{"input": "exclude NAs from data [CODESPLIT] data <- data[complete.cases(data$average), ] summary(data) nrow(data) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " 2. compare model indices (ICC, conditional and margina RÂ²)  [CODESPLIT] tab_model(model0.1, model1, model1_full) tab_model(model0.1, model2, model2_full) tab_model(model0.2, model3, model3_full) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot robust model 1  get predicted Values for different levels of rsa_socialresources [CODESPLIT] gg_model1 <- ggpredict(model1_robust, c(\"daycount[0, 25, 50, 75]\", \"rsa_socialresources[meansd]\", \"diagnosis\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "count of articles that generated data using experimental techniques (includes articles that use both, percentage calculated using total empirical articles)  [CODESPLIT] GenerateData[2,2] <- sum(MMCPSR_emp$EHPdata) GenerateData[2,3] <- sum(MMCPSR_emp$EHPdata)/nrow(MMCPSR_emp) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Data organization   Rename columns/variable names to make it simpler variable names [CODESPLIT] colnames(srmadata) srmadata <- srmadata %>% rename(assessors = 2, journal = 3, pubyear = 4, pubmonth = 5, study.title = 6, pmid = 7, regist = 8, regist.num = 9, protocol = 10, title.ident = 11, ab.sources = 12, ab.eleg.crit = 13, ab.particip = 14, ab.interv = 15, ab.effect = 16, ab.included = 17, ab.outcome = 18, in.picos = 19, me.database = 20, me.search.avai = 21, me.grey.lit = 22, me.date.just = 23, me.lang.num = 24, me.picos.desc = 25, me.sele.dup = 26, me.extr.dup = 27, me.rob.desc = 28, me.rob.dup = 29, me.stat.desc = 30, me.heterog = 31, item.removed01 = 32, re.flowdia = 33, re.ssizes = 34, re.picos.desc = 35, re.lengths = 36, re.estim.desc = 37, re.meta.studies = 38, re.rob = 39, re.deviations = 40, di.spin = 41, di.rob.studies = 42, di.limitations = 43, data.statem = 44, fund.statem = 45, funders = 46, coi.statem = 47) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Redefine columns types to factor [CODESPLIT] srmadata <- data.frame(srmadata) srmadata %>% mutate_if(is.character, as.factor) %>% str() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Add labels to variables in participant dataframe;; then, generate table  [CODESPLIT] table1::label(participant$ab.particip) <- \"Description of participants (ab)\" table1::label(participant$re.picos.desc) <- \"Detailed studies' characteristics\" table1::table1(~ab.particip + re.picos.desc, data = participant) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " HISTOGRAM   Generate a histogram of density of scores achieved by the 104 assessed studies  Create a 0 or 1 dataset, where No0 and Yes1, with respective study IDs (variable \"id\")  [CODESPLIT] id <- (1:104) allbinary <- data.frame(id, transparency, completenessbinary, participant, intervention, outcome, rigorbinary, appraisalbinary) allbinary[allbinary == \"Yes\"] <- \"1\" allbinary[allbinary == \"No\"] <- \"0\" lapply(allbinary,as.numeric) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Set the dataframe as numeric so that we can sum up recommended practices for each study (variable \"yes.score\")  [CODESPLIT] allbinary[] <- lapply(allbinary, function(x) as.numeric(as.character(x))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a new dataframe (binarytotalsdf) with a new variable (yes.score) that reflects the number of recommended practices from each study [CODESPLIT] binary-scoredf <- allbinary %>% mutate(yes.score = regist + protocol + me.search.avai + data.statem + + title.ident + + ab.sources + ab.eleg.crit + ab.included + in.picos + me.picos.desc + re.flowdia + + re.ssizes + re.lengths + fund.statem + coi.statem + ab.particip + re.picos.desc + + ab.interv + re.picos.desc + ab.outcome + me.stat.desc + me.heterog + re.estim.desc + + re.meta.studies + me.grey.lit + me.date.just + me.lang.num + me.sele.dup + me.extr.dup + + me.rob.desc + me.rob.dup +re.rob + re.deviations + di.spin + di.rob.studies + + di.limitations)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create the histogram [CODESPLIT] histplot.score <- ggplot(binary-scoredf, aes(x=yes.score)) + geom_histogram(binwidth=1, color=\"black\", fill=\"lightblue\") histplot.score + scale_x_continuous(name=\"Number of recommended practices (max: 36 items)\", breaks=seq(0,36,2)) + scale_y_continuous(name=\"Frequency of publications\", limits=c(0, 20)) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = \"black\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Load data: data file \"hfp\" contains individuallevel foraging return data (kcal) at the daily level along with the date, unique camp id, age, sex, and unique person id.  [CODESPLIT] hfp <- read.csv(\"food_pro_data.csv\", as.is=T) h <- hfp[hfp$age > 5,] h$age_z <- (h$age-mean(h$age, na.rm=T))/sd(h$age, na.rm=T) h$month <- month(ymd(h$date)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot the marginal/conditional effects for predicted values and probability of zerodays [CODESPLIT] p <- conditional_effects(hadza_lognormal_returns_model, effects=c(\"age_z:sex\"))  # includes the hurdle component p2 <- conditional_effects(hadza_lognormal_returns_model, effects=c(\"age_z:sex\"), dpar=\"hu\") # to see the hurdle component (Pr zero-days) plot(p)[[1]] + scale_x_continuous(breaks = c(-1, 0, 1, 2), labels= round(c(-1, 0, 1, 2)*sd(h$age, na.rm=T) + mean(h$age, na.rm=T))) + ggplot2::labs(x=\"Age\", y=\"kcal/day\") +ggplot2::lims(y=c(0,6500)) + theme_classic() plot(p2)[[1]] + ggplot2::lims(y=c(0,1)) + labs(x=\"Age\", y=\"Probability of zero day\\n(hurdle component of model)\") + theme_classic() + scale_x_continuous(breaks = c(-1, 0, 1, 2), labels= round(c(-1, 0, 1, 2)*sd(h$age, na.rm=T) + mean(h$age, na.rm=T))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate overall mean standard length (Ls) [CODESPLIT] Ls <- mean(standard.length, na.rm = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate Psi of an AR(1) model [CODESPLIT] Psi = diag(p) diag(Psi) = runif(p,b.ar.min,b.ar.max) return(Psi) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We use the R package qgraph to visualize our matrix [CODESPLIT] if (!require(qgraph)) install.packages(\"qgraph\");; require(qgraph) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "VISUALIZE YOUR NETWORKS  Make sure you have the right package: qgraph [CODESPLIT] if (!require(qgraph)) install.packages(\"qgraph\");; require(qgraph) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " load source functions and data  generates four data sets: FF, FS, SF, SS  [CODESPLIT] source(\"prep-Exp1-data.r\") source(\"pmwg-DIC.r\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "estimate model independently for each condition in the experiment [CODESPLIT] for(condition in names(all.data)) { cat(\"\\n\\n\\n\\nEstimating model for: \", condition, \"\\n\\n\") fnam <- paste0(\"Exp1-LBA-\", model.par, \"-\", condition, \".RData\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Then write loop to go through all Scopus IDs, adding the output to the first dataframe  NOTE: create 'output' folder in your working directory before running the following code  [CODESPLIT] for (i in 1:length(authors)){ tryCatch({ #using tryCatch () to go around error https://stackoverflow.com/questions/14748557/skipping-error-in-for-loop. So here it skips the authorIDs with an error and keeps going res = retrievalByAuthorID(authors[i], apik) M2 = res$M output <- rbind(M,M2) write.csv(output, paste0(output$AU_ID[1],\".csv\"), row.names=F) #creates seperate csv's for each Scopus ID }, error=function(e){}) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Select first letter of first name + surname [CODESPLIT] for(i in names(x)){ firstname = word(x[[i]],-1) initial = substring(firstname, 1, 1) surname = word(x[[i]],-2) surname[is.na(surname)] <- \"\" tot = as.data.frame(paste(initial, surname,sep = \".\")) tot[tot==\".\"]=\"\" x[[i]] <- tot  #join first and last name with a full stop } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Add a column per author, indicating which papers belong to them (0 or 1)  [CODESPLIT] authors = as.character(unique(authorswithAPInotworking$author.name)) x[is.na(x)]=\"\" for(i in (authors)){ x$i = rowSums(x == i) colnames(x)[colnames(x) == 'i'] <- i } ncol.new = ncol(x) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Have a look at distribution first and last years, to look at outliers  [CODESPLIT] hist(as.numeric(new$first.year)) hist(as.numeric(new$last.year))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add position of word in sentence (prefabricated list made in Python) [CODESPLIT] positionlist <- read.delim(\"U:/surfdriveRU/Thesis analyse/LMM analyse/positionlist.txt\", header = FALSE) eyetrackingdata$position <- positionlist$V1 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "density plotting function [CODESPLIT] denschart3 <- function (x, labels = NULL, groups = NULL, gdata = NULL, cex = par(\"cex\"), pt.cex = cex, bg = par(\"bg\"), color = \"grey20\", colorHPDI =\"grey60\", HPDI=0.9, vline = NULL, gcolor = par(\"fg\"), lcolor = \"gray\", xlim = range(unlist(x)), yvals = 1:length(x), yextra=0.7, main = NULL, xlab = NULL, ylab = NULL, height=0.7 , border=NA, adjust=1, ...) { opar <- par(\"mai\", \"mar\", \"cex\", \"yaxs\") on.exit(par(opar)) par(cex = cex, yaxs = \"i\") if (!is.list(x)) stop(\"'x' must be a list of vectors or matrices\") n <- length(x) glabels <- NULL if (is.list(x)) { if (is.null(labels)) labels <- names(x) if (is.null(labels)) labels <- as.character(1L:n) labels <- rep_len(labels, n) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "   Data wrangling    recode values for missing data (9, 999) in whole dataset as NA  [CODESPLIT] data <- data %>% mutate_all(~na_if(., -999)) data <- data %>% mutate_all(~na_if(., -9)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Moderated regression    center variables [CODESPLIT] data$ls.1c <- data$ls.1 - 3.57 data$sc.c  <- scale(data$sc,   center = T, scale = F) data$mob.c <- scale(data$mob,  center = T, scale = F) data$age.c <- scale(data$age,  center = T, scale = F) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get standardised regression coefficients [CODESPLIT] lm.beta(fit1) lm.beta(fit2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots Moderated regression [CODESPLIT] psych::describe(data[c(\"sc\", \"mob\")]) interact_plot(fit1, pred = mob.c, modx =sc.c, #partial.residuals = TRUE, interval = TRUE, int.width = 0.95, modx.values = c(-1.48, 0, 1.48), colors = c(\"#D9ED92\", \"#52B69A\", \"#1E6091\"), modx.labels = c(\"Low (- 1SD)\", \"Middle (Mean)\", \"High (+ 1SD)\"), legend.main = \"Perceived status loss\") + labs(y = \"Predicted life satisfaction\", x = \"Upward mobility beliefs\") + guides(fill = guide_legend(title = \"Perceived status loss\")) + coord_cartesian(ylim = c(1,7)) + scale_y_continuous(expand = c(0, 0), breaks = c(1,2,3,4,5,6,7)) + scale_x_continuous(expand = c(0.05, 0.05), breaks = c(-2.94,-1.94, -0.94, 0.06, 1.06, 2.06, 3.06), labels = c(0,1,2,3,4,5,6)) + theme_classic(base_size = 15) + theme( axis.title = element_text(colour = \"black\", size = 19, margin = margin(t = 0, r = 15, b = 0, l = 0)), axis.text.x  = element_text(colour = \"black\", size = 19, margin = margin(t = 15, r = 0, b = 0, l = 0)), axis.text.y  = element_text(colour = \"black\", size = 19), legend.position = c(0.7,0.85), legend.title = element_text(size = 17), legend.text = element_text(size = 17), legend.key.width = unit(1,\"cm\") ) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get diffs  get rownumbers in order to get the corresponding RecordingTime [CODESPLIT] rownums_parent <- fi_pairs$fiend[latencies[[curr_colname]][[curr_hitname]]$fi_pairs[[1]]] rownums_successor <- fi_pairs$fistart[latencies[[curr_colname]][[curr_hitname]]$fi_pairs[[2]]]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Funnel plot  carry out trimandfill analysis [CODESPLIT] taf<-trimfill(m.random,main=\"\", ma.fixed = FALSE, fixed = FALSE, random = TRUE, label=F) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "draw funnel plot with missing studies filled in [CODESPLIT] funnel(taf, legend=TRUE) summary(trimfill(m.random)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create container variable to store results [CODESPLIT] if(exists(\"data.full\") == 0){ subject <- c(1:50) mean.3SD <- NA sd.3SD <- NA min.3SD <- NA max.3SD <- NA container.3SD <- data.frame(subject, mean.3SD, sd.3SD, min.3SD, max.3SD) } container.3SD$mean.3SD[i] <- mean(data.3SD$rt) container.3SD$sd.3SD[i] <- sd(data.3SD$rt) container.3SD$min.3SD[i] <- container.3SD$mean.3SD[i] - 3*container.3SD$sd.3SD[i] container.3SD$max.3SD[i] <- container.3SD$mean.3SD[i] + 3*container.3SD$sd.3SD[i] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "setting up colors and line types for plots [CODESPLIT] ns=unique(qms$name) nl=length(ns) disp=data.frame(name=ns,lty=rep(1:9,nl)[1:nl]) disp$col=ifelse(grepl('Hit|Rod',disp$name),'#880101','#aaaaaa') disp$lty[grepl('Hit|Rod',disp$name)]=1:2 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot the results of the vector fits  color ages > 85 in red [CODESPLIT] col <- c(\"gray\", \"red\")[1 + (age >= 85)] main <- \"PCoA of Pollen: pollen and independent vectors fitted onto ordination\" plot(poln.ord$points, pch = 21, type = \"p\", xlab = paste0(\"PCoA I (\", round(var.expl[1] * 100, 1), \"%)\"), ylab = paste0(\"PCoA II (\", round(var.expl[2] * 100, 1), \"%)\"), main = main, col = \"black\", bg = col, cex = 1.5, xaxt = \"n\", yaxt = \"n\", bty = \"n\") axis(1, at = seq(-0.5, 1, by = 0.25)) axis(2, at = seq(-1, 0.25, by = 0.25)) legend(0.5, -0.85, pch = c(21, 21, NA), lty = c(NA, NA, 1), lwd = c(NA, NA, 2), col = c(\"black\", \"black\", \"light blue\"), pt.bg =  c(\"gray\",\"red\", NA), cex = 1.25, legend = c(\"Post 85k\", \"Pre 85k\", \"Lake Level\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create categorical variables [CODESPLIT] share <- to_factor( share, select = c(\"health_past3months\", \"wave\", \"gender\", \"covid_affected\", \"partnerinhh\", \"covid_regime_si3\", \"covid_regime_ch3\") ) share$stringency_index <- share$global_covid_regime_si3 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Overall mean of Social factors [CODESPLIT] share_all <- data_filter(share, !is.na(health_past3months)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Model comparison   apply hierarchical versions of CC, EWA, LMM, and motivational EWA models  [CODESPLIT] setwd(\"SET TO MODEL FILE DIRECTORY\") data <- list(\"groupSize\", \"ngroups\", \"ntrials\", \"ntokens\", \"pi\", \"vals\",\"c\",\"Gc\",\"c_choice_index\",\"Ga\") #data inputted into jags params <- c(\"mu_c\") #parameters we'll track in jags samplesCC <- jags(data, inits=NULL, params, model.file =\"CC_group.txt\", n.chains=3, n.iter=5000, n.burnin=1000, n.thin=1) params <- c(\"mu_c\") #parameters we'll track in jags samplesEWA <- jags(data, inits=NULL, params, model.file =\"EWA_group.txt\", n.chains=3, n.iter=5000, n.burnin=1000, n.thin=1) params <- c(\"nu_c\") #parameters we'll track in jags samplesLMM <- jags(data, inits=NULL, params, model.file =\"LMM_group.txt\", n.chains=3, n.iter=5000, n.burnin=1000, n.thin=1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number of subjects in each group  Create a group variable according to each cluster condition [CODESPLIT] if (size == 1){ N.size = N/K n.group = unlist(lapply(1:K, function(k) rep(k,N.size))) } if (size == 2){ if (K == 2){ N.1 = 0.10*N N.2 = N - N.1 n.group = c(rep(1,N.1),rep(2,N.2)) } if (K == 4){ N.1 = 0.10*N N.rest = N - N.1 N.size = N.rest/(K-1) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size),rep(4,N.size)) }} if (size == 3){ if (K == 2){ N.1 = 0.6*N N.2 = N - N.1 n.group = c(rep(1,N.1),rep(2,N.2)) } if (K == 4){ if (N == 20){ N.1 = 0.6*N N.rest = N - N.1 N.size = floor(N.rest/(K-1)) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size+1),rep(4,N.size+1)) } else{ N.1 = 0.6*N N.rest = N - N.1 N.size = N.rest/(K-1) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size),rep(4,N.size)) }}} ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create variable subjno  Scramble data so people appear unordered in the final dataset [CODESPLIT] ID = expand.grid(1:T,1:N)[,2] data = cbind(data,ID) p = ncol(Psi[[1]]) Y = matrix(0,nrow(data),p) colnames(Y) = sprintf(\"Y%d\",seq(1:p)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Frequency plot [CODESPLIT] top.freqs = sort(colSums(data.table), decreasing = T)[1:40] cex.val = 1.6 cairo_pdf('pdfs/top_freq_segments.pdf', width = 16, height = 10) par(family = \"Doulos SIL\") plot(top.freqs ~ seq_along(top.freqs), xlab = 'Frequency rank', ylab = 'Frequency', xlim = c(1, 40), type = 'n', xaxt = 'n', cex.lab = cex.val, cex.axis = cex.val) axis(1, at = seq_along(top.freqs), labels = seq_along(top.freqs), cex.axis = 1.1) lines(seq_along(top.freqs), top.freqs, lty = 2, col = 'grey') text(seq_along(top.freqs), top.freqs, labels = names(top.freqs), cex = 2) dev.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Contrast between countries as a new variable [CODESPLIT] d$CountC<-as.numeric(as.integer(as.factor(d$Country))-1.5) table(d$Country,d$CountC) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Estimated Marginal Means Model 3 [CODESPLIT] m1 <- ggemmeans(model3, c(\"welle [1:4 by=.2]\", \"isced\")) m2 <- ggemmeans(model3, c(\"welle [1:4 by=.2]\", \"aee_oecd_between_z2\")) m3 <- ggemmeans(model3, c(\"welle [1:4 by=.2]\", \"lone6_between_z2\")) m1$Model = \"ISCED\" levels(m1$group) <- c(\"low\", \"middle\", \"high\") m2$Model = \"Income\" levels(m2$group) <- c(\"-1 SD\", \"Mean\", \"+1 SD\") m3$Model = \"Loneliness\" levels(m3$group) <- c(\"-1 SD\", \"Mean\", \"+1 SD\") create_plot(m1, title = \"(a) PF and Education\") create_plot(m2, title = \"(b) PF and Income\") create_plot(m3, title = \"(c) PF and Loneliness\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Matrix with correlations and pvalues [CODESPLIT] Cormatrix1 <- data.frame(matrix(NA,nrow = 15, ncol = 15)) for(i in 1:14){ for(j in (i+1):15){ Cormatrix1[i,j] <- round(cor(Cordata[,i], Cordata[,j]), digits=2) Cormatrix1[j,i] <- round(cor.test(Cordata[,i], Cordata[,j])$p.value, digits=2) } } Cormatrix1 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Matrix with correlations and significance stars [CODESPLIT] Cormatrix2 <- data.frame(matrix(NA,nrow = 15, ncol = 15)) for(i in 1:14){ for(j in (i+1):15){ Cormatrix2[i,j] <- round(cor(Cordata[,i], Cordata[,j]), digits=2) Cormatrix2[j,i] <- ifelse(cor.test(Cordata[,i], Cordata[,j])$p.value<=0.05, \"*\", \"no\") } } Cormatrix2 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate Psi matrix for each cluster [CODESPLIT] for (n in 1:length(N)){ for (t in 1:length(T)){ for (p in 1:length(P)){ for (k in 1:length(K)){ for (d in 1:length(diff)){ for (r in 1:R){ Psi.list = lapply(1:K[k], function(kl) Psi.Matrix.Diff(P[p],b.ar.min,b.ar.max,b.cr.min,b.cr.max,d)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Load training and testing set [CODESPLIT] load(file = paste(\"Data_Block_Cluster_MVAR_RE_N_\",N[n],\"_T_\",T[t],\"_P_\",P[p],\"_K_\",K[k], \"_Diff_\",diff[d],\"_size_\",size[s],\"_R_\",r,\".RData\",sep = \"\")) MSE.MVAR.PS.Sys = MSE.Sys(data.list,P[p]) save(MSE.MVAR.PS.Sys, file = paste(\"MSE_MVAR_Cluster_MVAR_RE_N_\",N[n],\"_T_\",T[t],\"_P_\",P[p],\"_K_\",K[k], \"_Diff_\",diff[d],\"_size_\",size[s],\"_R_\",r,\".RData\",sep = \"\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " standardize the model inputs, excluding the response and random effects  [CODESPLIT] d_std <- stand(d, cols = f2) # use the fitting function for convenience ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " > Left panel: QQplot (uniform distribution)  > Right panel: Residuals against predicted values;; shaded (due to sample size) with extreme residuals colored red, and   3) MAIN EFFECTS OF KEY VARIABLES   FormatInfo  [CODESPLIT] emmip(CorResult, ~FormatInfo, type = \"response\", CIs = TRUE) (emm <- emmeans(CorResult, specs = ~FormatInfo, type = \"response\")) pairs(emm) EffPlotData_CorrFormat <- summary(emm)  ## Data for Fig. 3 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Rescale order of variables on yaxis (for BRT figures) [CODESPLIT] BRT.plot.label.limits <- c(\"site.centrality\", \"mean.annual.flow\", \"basin.area\", \"site.long\", \"site.lat\", \"pct.ISC\", \"pct.urb\", \"pct.ag\", \"pct.for\", \"ALG.cover\", \"NAT.cover\", \"LWD.reach\", \"DOC\", \"cond\", \"pH.lab\", \"total.P\", \"NH4\", \"NO3\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Rename variables names on yaxis (for BRT figures) [CODESPLIT] BRT.plot.labels <- c(\"site.centrality\" = \"Cent\", \"mean.annual.flow\" = \"Flow\", \"basin.area\" = \"Area\", \"site.long\" = \"Long\", \"site.lat\" = \"Lat\", \"pct.ISC\" = \"ISC\", \"pct.urb\" = \"Urb\", \"pct.ag\" = \"Ag\", \"pct.for\" = \"For\", \"ALG.cover\" = \"Alg\", \"NAT.cover\" = \"Nat\", \"LWD.reach\" = \"LWD\", \"DOC\" = \"DOC\", \"cond\" = \"Cond\", \"pH.lab\" = \"pH\", \"total.P\" = \"TP\", \"NH4\" = expression(NH[4]), \"NO3\" = expression(NO[3])) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Sample posetrior and prior for graphical comparison [CODESPLIT] post1<-extract.samples(m1) set.seed(42) prio1<-extract.prior(m1,n=10000) save.image(file=\"posterior_samples_single.RData\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlation matrix for the DVs [CODESPLIT] judcorMat2 <- lowerCor(judgmentRatings2) corr.test(judgmentRatings2) corrplot(judcorMat2, method=\"color\", type = 'lower',tl.col=\"black\", addCoef.col = \"black\", tl.srt = 45) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Subset data that includes only partner cooperation means within 0.4  0.6 [CODESPLIT] bound_data <- subset(expt1_data, expt1_data$cooplc_means >= 0.4 & expt1_data$cooplc_means <= 0.6 & expt1_data$readlc_means >= 0.4 & expt1_data$readlc_means <= 0.6) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot histogram of chances that partner's choice was positive for all data [CODESPLIT] coop_hist_ggplot <- ggplot(all_pc, aes(x = alllc_means * 100)) + geom_histogram(aes(fill = included), bins = 50) +  # plot histogram scale_fill_manual(values = c(\"black\", \"grey50\"), name=\"\", label=c(\"Included\", \"Not included\")) +  # color values inside and outside of 0.4-0.6 differently labs(x = \"Percent partner positive actions\", y = \"Number of participants\") + # label axes theme_classic() + # use classic theme theme(axis.title=element_text(size=45), axis.text=element_text(size=30), legend.text=element_text(size=30), legend.position = c(0.25, 0.9), legend.key.size = unit(2.5, 'lines')) png(file = \"figures/partner_action_histogram.png\", width = 1200, height = 750) # open device plot(coop_hist_ggplot) # plot figure dev.off() # close device ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  Analyze accuracy as a function of payoff scheme (Standard or Costly), context (Cooperation or Newspaper), and partner action (Cooperate or Defect)   Conduct binomial GLMM of payoff scheme * partner action + context for memory accuracy  [CODESPLIT] accuracy_glmer_full <- glmer(accuracy ~ payoff_scheme * partner_action * context + (1 | subject), bound_data, family = binomial(link = \"logit\"))   # calculate GLMM of full model ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Conduct binomial GLMMs for memory accuracy to calculate BIC values to transform to Bayes factors [CODESPLIT] accuracy_glmer_null <- summary(glmer(accuracy ~ (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for null model accuracy_glmer_payoff <- summary(glmer(accuracy ~ payoff_scheme  + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff scheme accuracy_glmer_context <- summary(glmer(accuracy ~ context + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for context accuracy_glmer_action <- summary(glmer(accuracy ~ partner_action + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for partner action accuracy_glmer_payoff_action <- summary(glmer(accuracy ~ payoff_scheme + partner_action + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff_scheme + partner_action accuracy_glmer_payoff_action_inter <- summary(glmer(accuracy ~ payoff_scheme * partner_action + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff_scheme * partner_action accuracy_glmer_action_context <- summary(glmer(accuracy ~ partner_action + context + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for partner_action + context accuracy_glmer_action_context_inter <- summary(glmer(accuracy ~ partner_action * context + (1 | subject), bound_data, family = binomial(link = \"logit\")))    # calculate GLMM for partner_action * context accuracy_glmer_payoff_action_context <- summary(glmer(accuracy ~ payoff_scheme + partner_action + context + (1 | subject), bound_data, family = binomial(link = \"logit\")))   # calculate GLMM for payoff_scheme * partner_action * context last_accuracy_glmer_null <- summary(glmer(accuracy ~ (1 | subject), last_data, family = binomial(link = \"logit\")))    # calculate GLMM for null model last_accuracy_glmer_payoff <- summary(glmer(accuracy ~ payoff_scheme  + (1 | subject), last_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff scheme last_accuracy_glmer_action <- summary(glmer(accuracy ~ partner_action + (1 | subject), last_data, family = binomial(link = \"logit\")))    # calculate GLMM for partner action last_accuracy_glmer_payoff_action <- summary(glmer(accuracy ~ payoff_scheme + partner_action + (1 | subject), last_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff_scheme + partner_action last_accuracy_glmer_payoff_action_inter <- summary(glmer(accuracy ~ payoff_scheme * partner_action + (1 | subject), last_data, family = binomial(link = \"logit\")))    # calculate GLMM for payoff_scheme * partner_action ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Conduct binomial GLMM of payoff scheme * partner action + context for memory accuracy [CODESPLIT] last_accuracy_glmer_full <- glmer(accuracy ~ payoff_scheme * partner_action + (1 | subject), last_data, family = binomial(link = \"logit\"))   # calculate GLMM of full model ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate correlation between mean number of contacts and mean memory accuracy [CODESPLIT] coop_contacts_cor2 <- cor.test(expt2_data_subj$accuracy, expt2_data_subj$contacts)  # calculate network size/accuracy correlation coop_contacts2_bfdf <- data.frame(accuracy = expt2_data_subj$accuracy, contacts = expt2_data_subj$contacts) # create new data frame for Bayesian analysis coop_contacts2_lmbf <- lmBF(accuracy ~ contacts, data = coop_contacts2_bfdf)  # calculate Bayes regression coop_contacts2_bf <- extractBF(coop_contacts2_lmbf)$bf  # extract Bayes factor ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Create dummy variable for ethnicity: 0  option 4 only (Anglo/White), 1  any other option / combination of options  [CODESPLIT] table(survey$ethnic_group[!duplicated(survey$id)], useNA = \"ifany\") survey$ethnicity <- ifelse(survey$ethnic_group == \"4\", 0, 1) table(survey$ethnicity[!duplicated(survey$id)], useNA = \"ifany\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Create dummy variable for SES: 1  mother or father completed at least some college (4  some college), 0  otherwise  [CODESPLIT] table(survey$mother_educationlevel[!duplicated(survey$id)], useNA = \"ifany\") table(survey$father_educationlevel[!duplicated(survey$id)], useNA = \"ifany\") survey$SES <- ifelse(survey$mother_educationlevel >= 4 | survey$father_educationlevel >= 4, 1, 0) table(survey$SES[!duplicated(survey$id)], useNA = \"ifany\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Weekend  Create dummy variable for weekend: 1  weekend, 0  weekday  [CODESPLIT] survey$weekend <- ifelse(weekdays(survey$StartDate, abbr = TRUE) %in% c(\"Sat\", \"Sun\"), 1, 0) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create dummy variable for mixed interaction partners [CODESPLIT] survey$no_partners <- apply(survey[c(\"close_peers\", \"family\", \"weak_ties\")], 1, sum) survey$mixed_partner <- ifelse(survey$no_partners > 1, 1, 0) table(survey$mixed_partner, useNA = \"ifany\") # number of observations with mixed interaction partners: 11479 (S1) / 5797 (S2) table(apply(survey[c(\"close_peers\", \"family\", \"weak_ties\")], 1, sum), useNA = \"ifany\") survey$close_peers_all <- survey$close_peers survey$family_all <- survey$family survey$weak_ties_all <- survey$weak_ties survey$close_peers <- ifelse(survey$mixed_partner == 1, NA, survey$close_peers) survey$family <- ifelse(survey$mixed_partner == 1, NA, survey$family) survey$weak_ties <- ifelse(survey$mixed_partner == 1, NA, survey$weak_ties) table(survey$close_peers, useNA = \"ifany\") # number of observations with interactions with close peers ONLY: 17194 (S1) / 9406 (S2) table(survey$interacting_people[survey$close_peers == 1]) table(survey$family, useNA = \"ifany\") # number of observations with interactions with family ONLY: 2450 (S1) / 1946 (S2) table(survey$interacting_people[survey$family == 1]) table(survey$weak_ties, useNA = \"ifany\") # number of observations with interactions with weak ties ONLY: 4336 (S1) / 2823 (S2) table(survey$interacting_people[survey$weak_ties == 1]) table(apply(survey[c(\"close_peers\", \"family\", \"weak_ties\")], 1, sum), useNA = \"ifany\") # 23980 (S1) / 14175 (S2) length(which(survey$mixed_mode == 1 | survey$SKIP2 == 1 | survey$OTHER2 == 1 | is.na(survey$interacting_people) | survey$mixed_partner == 1)) table(apply(survey[c(\"close_peers_all\", \"family_all\", \"weak_ties_all\")], 1, sum), useNA = \"ifany\") # 23980, 10176, 1303 (S1) / 14175, 5116, 681 (S2) length(which(survey$mixed_mode == 1 | survey$SKIP2 == 1 | survey$OTHER2 == 1 | is.na(survey$interacting_people))) survey$no_partners2 <- apply(survey[c(\"friends_roommates\", \"significant_other\", \"family_all\", \"weak_ties_all\")], 1, sum) survey$mixed_partner2 <- ifelse(survey$no_partners2 > 1, 1, 0) table(survey$mixed_partner2, useNA = \"ifany\") # number of observations with mixed interaction partners: 13456 (S1) / 6847 (S2) table(apply(survey[c(\"friends_roommates\", \"significant_other\", \"family_all\", \"weak_ties_all\")], 1, sum), useNA = \"ifany\") survey$friends_roommates2 <- ifelse(survey$mixed_partner2 == 1, NA, survey$friends_roommates) survey$significant_other2 <- ifelse(survey$mixed_partner2 == 1, NA, survey$significant_other) survey$family2 <- ifelse(survey$mixed_partner2 == 1, NA, survey$family_all) survey$weak_ties2 <- ifelse(survey$mixed_partner2 == 1, NA, survey$weak_ties_all) table(survey$friends_roommates2, useNA = \"ifany\") # number of observations with interactions with friends and roommates ONLY: 12882 (S1) / 6865 (S2) table(survey$interacting_people[survey$friends_roommates2 == 1]) table(survey$significant_other2, useNA = \"ifany\") # number of observations with interactions with significant others ONLY: 2335 (S1) / 1491 (S2) table(survey$interacting_people[survey$significant_other2 == 1]) table(survey$family2, useNA = \"ifany\") # number of observations with interactions with family ONLY: 2450 (S1) / 1946 (S2) table(survey$interacting_people[survey$family2 == 1]) table(survey$weak_ties2, useNA = \"ifany\") # number of observations with interactions with weak ties ONLY: 4336 (S1) / 2823 (S2) table(survey$interacting_people[survey$weak_ties2 == 1]) table(apply(survey[c(\"friends_roommates2\", \"significant_other2\", \"family2\", \"weak_ties2\")], 1, sum), useNA = \"ifany\") # 22003 (S1) / 13125 (S2) length(which(survey$mixed_mode == 1 | survey$SKIP2 == 1 | survey$OTHER2 == 1 | is.na(survey$interacting_people) | survey$mixed_partner2 == 1)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create dummy variable for significant others [CODESPLIT] survey$significant_other <- ifelse(survey$mixed_mode == 1 | survey$SKIP2 == 1 | survey$OTHER2 == 1 | is.na(survey$interacting_people), NA, ifelse(survey$Significant_other == 1, 1, 0)) table(survey$significant_other, useNA = \"ifany\") # number of observations with interactions with significant others: 6681 (S1) / 3822 (S2) table(survey$interacting_people[survey$significant_other == 1]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "converting PPId and Statement to factors [CODESPLIT] Data$PPID <- as.factor(Data$PPID) Data$Statement <-as.factor(Data$Statement)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "find Q1, Q3, and interquartile range for values in column A  [CODESPLIT] Q1 <- quantile(DV2, .25) Q3 <- quantile(DV2, .75) IQR2 <- IQR(DV2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "only keep rows in dataframe that have values within 1.5*IQR of Q1 and Q3 [CODESPLIT] no_outliers2 <- subset(df, DV2> (Q1 - 1.5*IQR2) & DV2< (Q3 + 1.5*IQR2)) no_outliers3 <- subset(df, DV3> (Q1 - 1.5*IQR3) & DV3< (Q3 + 1.5*IQR3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot beta weights for interaction [CODESPLIT] betalms #  called \"beta\" but represents \"b\" betalms <- betalms[,1:2] names(betalms) <- c(\"b\", \"SE\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot beta weights for engagement [CODESPLIT] betalms_eng betalms_eng <- betalms_eng[,1:2] names(betalms_eng) <- c(\"b\", \"SE\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Arrange b coeff plots [CODESPLIT] grid.arrange(betalms_ope_p, betalms_eng_p, betalms_p, ncol = 3, nrow = 1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot slopes [CODESPLIT] listofcharts = list() # create empty list for charts index = 0 # zero the index for (df in listofdfs) { index = index + 1 xlab_str = paste0(\"Economic beliefs in \", names(listofdfs)[index]) listofcharts[[index]] <- ggpredict(listoflms[[index]], terms = c(\"conservation2_s_c\", \"polit_eng_c[-0.14,0.14]\"), type = \"fe\") %>% plot(colors = \"bw\") + ggtitle(names(listoflms[index])) + xlab(\"NSC\") + ylab(\"Economic beliefs\") + labs(linetype = \"Political \\nengagement\") + scale_linetype_manual(values=c(\"solid\", \"dashed\"), labels = c(\"Low\", \"High\")) + theme_classic() + theme(legend.position = \"none\") + ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Does the start model fit the data significantly better as compared to a model without random intercepts over items? [CODESPLIT] tic();; start_min_item_intercepts <- glmer(bin_score ~ input*testmoment*learningtype + (1|participant), family = 'binomial', data = data, control = glmerControl(optimizer = \"bobyqa\", optCtrl=list(maxfun=1e5)));; toc() anova(start_min_item_intercepts, start) # Start model is significantly better (p < .001);; AIC difference of +- 240 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Random slope of input over item [CODESPLIT] tic();; input_item <- glmer(bin_score ~ input*testmoment*learningtype + (1+input|item) + (1|participant), family = 'binomial', data = data, control = glmerControl(optimizer = \"bobyqa\", optCtrl=list(maxfun=1e5)));; toc() anova(start, input_item) # Significant improvement (p = .03);; small AIC difference (3). summary(rePCA(input_item)) # All dimensions are supported by the data. >> KEEP ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Investigate model fit  Inspect residuals with a binned residual plot [CODESPLIT] residualsplot <- binnedplot(fitted(final), resid(final, type = \"response\"), cex.pts=1, col.int=\"black\", xlab = \"Estimated score (as probability)\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "model 1 is just baseHIne PHQ_9 severity in a simple logistic regression [CODESPLIT] HI_prognostic_model1 = glm(y ~ PHQ9_first, data = X_HI_only, family = \"binomial\") summary(HI_prognostic_model1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "model 3 is a logistic regression built using elastic net regularization  use elastic net to build this model [CODESPLIT] set.seed(12345678) fit_HI = glmnet(data.matrix(X_HI_only_forEN), y_HI_only, family=\"binomial\", alpha=.5) plot(fit_HI,label=TRUE) HI_prognostic_model3 = cv.glmnet(data.matrix(X_HI_only_forEN), y_HI_only, family=\"binomial\", alpha=.5) plot(HI_prognostic_model3) HI_prognostic_model3$lambda.min HI_prognostic_model3$lambda.1se print(coef(HI_prognostic_model3, s = \"lambda.min\")) print(coef(HI_prognostic_model3, s = \"lambda.1se\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate deviance statistic  logHIkeHIhood : sum from i1 to N of [ Yi*ln(P(Yi))+(1Yi)*ln(1P(yi))]  calculate brier score  brier_score: (1/n)*sum(pioi)^2 [CODESPLIT] log_likelihood_calculator_HI_1 = rep(NA,dim(X_HI_only_hold_out)[1]) brier_score_calculator_HI_1 = rep(NA,dim(X_HI_only_hold_out)[1]) log_likelihood_calculator_HI_2 = rep(NA,dim(X_HI_only_hold_out)[1]) brier_score_calculator_HI_2 = rep(NA,dim(X_HI_only_hold_out)[1]) log_likelihood_calculator_HI_3 = rep(NA,dim(X_HI_only_hold_out)[1]) brier_score_calculator_HI_3 = rep(NA,dim(X_HI_only_hold_out)[1]) log_likelihood_calculator_HI_4 = rep(NA,dim(X_HI_only_hold_out)[1]) brier_score_calculator_HI_4 = rep(NA,dim(X_HI_only_hold_out)[1]) log_likelihood_calculator_HI_5 = rep(NA,dim(X_HI_only_hold_out)[1]) brier_score_calculator_HI_5 = rep(NA,dim(X_HI_only_hold_out)[1]) for (i in 1:dim(X_HI_only_hold_out)[1]){ log_likelihood_calculator_HI_1[i] = y_HI_only_hold_out[i]*log(HI_prognosis_1[i])+(1-y_HI_only_hold_out[i])*log(1-HI_prognosis_1[i]) brier_score_calculator_HI_1[i] = (HI_prognosis_1[i]-y_HI_only_hold_out[i])^2 log_likelihood_calculator_HI_2[i] = y_HI_only_hold_out[i]*log(HI_prognosis_2[i])+(1-y_HI_only_hold_out[i])*log(1-HI_prognosis_2[i]) brier_score_calculator_HI_2[i] = (HI_prognosis_2[i]-y_HI_only_hold_out[i])^2 log_likelihood_calculator_HI_3[i] = y_HI_only_hold_out[i]*log(HI_prognosis_3[i])+(1-y_HI_only_hold_out[i])*log(1-HI_prognosis_3[i]) brier_score_calculator_HI_3[i] = (HI_prognosis_3[i]-y_HI_only_hold_out[i])^2 log_likelihood_calculator_HI_4[i] = y_HI_only_hold_out[i]*log(HI_prognosis_4[i])+(1-y_HI_only_hold_out[i])*log(1-HI_prognosis_4[i]) brier_score_calculator_HI_4[i] = (HI_prognosis_4[i]-y_HI_only_hold_out[i])^2 log_likelihood_calculator_HI_5[i] = y_HI_only_hold_out[i]*log(HI_prognosis_5[i])+(1-y_HI_only_hold_out[i])*log(1-HI_prognosis_5[i]) brier_score_calculator_HI_5[i] = (HI_prognosis_5[i]-y_HI_only_hold_out[i])^2 } deviance_for_HI_model_1 = sum(log_likelihood_calculator_HI_1,na.rm=TRUE) brier_score_for_HI_model_1 = (1/dim(X_HI_only_hold_out)[1])*sum(brier_score_calculator_HI_1,na.rm=TRUE) deviance_for_HI_model_2 = sum(log_likelihood_calculator_HI_2,na.rm=TRUE) brier_score_for_HI_model_2 = (1/dim(X_HI_only_hold_out)[1])*sum(brier_score_calculator_HI_2,na.rm=TRUE) deviance_for_HI_model_3 = sum(log_likelihood_calculator_HI_3,na.rm=TRUE) brier_score_for_HI_model_3 = (1/dim(X_HI_only_hold_out)[1])*sum(brier_score_calculator_HI_3,na.rm=TRUE) deviance_for_HI_model_4 = sum(log_likelihood_calculator_HI_4,na.rm=TRUE) brier_score_for_HI_model_4 = (1/dim(X_HI_only_hold_out)[1])*sum(brier_score_calculator_HI_4,na.rm=TRUE) deviance_for_HI_model_5 = sum(log_likelihood_calculator_HI_5,na.rm=TRUE) brier_score_for_HI_model_5 = (1/dim(X_HI_only_hold_out)[1])*sum(brier_score_calculator_HI_5,na.rm=TRUE) print(round(deviance_for_HI_model_1,1)) print(round(brier_score_for_HI_model_1,3)) print(round(deviance_for_HI_model_2,1)) print(round(brier_score_for_HI_model_2,3)) print(round(deviance_for_HI_model_3,1)) print(round(brier_score_for_HI_model_3,3)) print(round(deviance_for_HI_model_4,1)) print(round(brier_score_for_HI_model_4,3)) print(round(deviance_for_HI_model_5,1)) print(round(brier_score_for_HI_model_5,3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "here we create a PAIstyle model that includes variables and their interaction with tx  we use this model as a demonstration: [CODESPLIT] differential_model_6 = glm(y ~ tx*(PHQ9_first+WSAS_first+Employment_binary+Ethnicity_binary+GAD7_first+Phobia_Q3_first), data = X_training, family = \"binomial\") summary(differential_model_6) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "grab the indices of which individuals got tx0 (li) and tx1 (hi) [CODESPLIT] tx_HI_i = which(X_hold_out$tx==1, arr.ind = TRUE) tx_LI_i = which(X_hold_out$tx==0, arr.ind = TRUE) step_size = 150 window_size = 300 bin_number = ceiling((dim(X_hold_out)[1]-window_size)/step_size) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "the below command can be used to see what y limits and x limits should be used to standardize all plots [CODESPLIT] cat(\"y_limits\",round(c(min(observed_differential_response),max(observed_differential_response)),3),\"\\n\") cat(\"x_limits\",round(c(min(predicted_differential_response),max(predicted_differential_response)),3),\"\\n\") cat(\"predicted avg differential response (full sample avg):\",round(mean(differential_prediction),3),\"\\n\") cat(\"predicted avg differential response (binned avg):\",round(mean(predicted_differential_response),3),\"\\n\") predicted_avg_diff = mean(differential_prediction) plot(predicted_differential_response,observed_differential_response,'p',ylim=c(-.1,.3), xlim = c(-.1,.3),xlab=x_label) plot(predicted_differential_response,observed_differential_response,'p',ylim=c(-.1,.3), xlim = c(min(predicted_differential_response),max(predicted_differential_response)),xlab=x_label) cor_windows = cor(1:bin_number,observed_differential_response) cor_pred_diff = cor(predicted_differential_response,observed_differential_response) cat(\"windows correlation = \",round(cor_windows,3),\"\\n\") cat(\"predicted difference correlation = \",round(cor_pred_diff,3),\"\\n\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "we can use the tstatistic to adjust for the error around the slope  tvalue is the coefficient divided by its standard error [CODESPLIT] cat(\"slope t-stat = \",round(sc_coefs_summary$t.value[2],3),\"\\n\") print(round(sc_coefs_summary,3)) cat(\"\\n\") observed_range = max(observed_differential_response) - min(observed_differential_response) predicted_range = max(predicted_differential_response) - min(predicted_differential_response) tstat = sc_coefs_summary$t.value[2] slope = sc_coefs_summary$Estimate[2] model_evaluations[k,] = c(tstat, slope, observed_range, predicted_range, predicted_avg_diff, cor_pred_diff) model_predictions[k,] = predicted_differential_response model_results[k,] = observed_differential_response } print(round(model_evaluations,3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "comparisons to average for each trait [CODESPLIT] summary(lmer(RATINGc ~ TSELFc* EX + SMEANc* EX + SDMEAN* EX +PSELFc* EX + (TSELFc + SMEANc + SDMEANc +PSELFc    | PID),data= subset(fmimlm, fmimlm$motive == 0 ) )) summary(lmer(RATINGc ~ TSELFc* OP + SMEANc* OP + SDMEAN* OP +PSELFc*OP+ (TSELFc + SMEANc + SDMEANc  +PSELFc   | PID),data= subset(fmimlm, fmimlm$motive == 0   ) )) summary(lmer(RATINGc ~ TSELFc* AG + SMEANc* AG + SDMEAN* AG +PSELFc*AG + (TSELFc + SMEANc + SDMEANc   +PSELFc  | PID),data= subset(fmimlm, fmimlm$motive == 0 ) )) summary(lmer(RATINGc ~ TSELFc* NE + SMEANc* NE + SDMEAN* NE +PSELFc*NE+ (TSELFc + SMEANc + SDMEANc  +PSELFc   | PID),data= subset(fmimlm, fmimlm$motive == 0  ) )) summary(lmer(RATINGc ~ TSELFc* CO + SMEANc* CO + SDMEAN* CO+PSELFc*CO+ (TSELFc + SMEANc + SDMEANc  +PSELFc   | PID), data= subset(fmimlm,fmimlm$motive == 0  ) )) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Conduct RSA  We want to estimate the RSA model in which the intercept is allowed to vary across states.  The equation for individual i living in state j (z  selfesteem, x  IV, y  SV) is:  zij  b0 + b1*xij + b2*yj + b3*xij^2 + b4*xij*yj + b5*yj^2 + uj + eij  specify and estimate this model:  [CODESPLIT] m.c <- lmer(selfesteem ~ IV.c + SV.c + IV2.c + IVSV.c + SV2.c + (1 | state), data = df) summary(m.c) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot the average surface using MLRSA_AverageSurfacePlot: [CODESPLIT] MLRSA_AverageSurfacePlot(m.c, name_vars=c(\"IV.c\",\"SV.c\",\"IV2.c\",\"IVSV.c\",\"SV2.c\"), outcome=\"selfesteem\", data=df, xlab=\"Individual-level values\", ylab=\"State-level values\", zlab=\"Self-esteem\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "only keeping latest reported dates for each patient [CODESPLIT] data_lab_res_dcr <- data_lab[!duplicated(data_lab$shcsid,fromLast=TRUE),] data_drug_res_dcr <- data_drug[!duplicated(data_drug$shcsid,fromLast=TRUE),] data_dis_res_dcr <- data_dis[order(data_dis$shcsid,data_dis$newdate),] data_dis_res_dcr <- data_dis_res_dcr[!duplicated(data_dis_res_dcr$shcsid,fromLast=TRUE),] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots with H(A|M)  nonnormalized conditional entropy Getting mean, median and standard deviation across different motifs per period  [CODESPLIT] N <- aggregate(CEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = length) MEAN <- aggregate(CEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = mean) MEDIAN <- aggregate(CEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = median) SD <- aggregate(CEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = sd) resultsmotifs_summary <- cbind.data.frame(N, MEAN$CEAuthoritiesMotifs, MEDIAN$CEAuthoritiesMotifs, SD$CEAuthoritiesMotifs) colnames(resultsmotifs_summary) <- c(\"DATE\",\"N\",\"MEAN\",\"MEDIAN\",\"SD\") resultsmotifs_summary$SE <- resultsmotifs_summary$SD / sqrt(resultsmotifs_summary$N) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots with H(A|M)/H(A)  normalized conditional entropy [CODESPLIT] resultsmotifs <- as.data.frame(rbind(results330,results350,results370,results390,results405,results415,results425, results435,results455,results470,results490,results600)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Getting mean, median and standard deviation across different motifs per period  [CODESPLIT] N <- aggregate(NormCEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = length) MEAN <- aggregate(NormCEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = mean) MEDIAN <- aggregate(NormCEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = median) SD <- aggregate(NormCEAuthoritiesMotifs ~ DATE, data = resultsmotifs, FUN = sd) Nresultsmotifs_summary <- cbind.data.frame(N, MEAN$NormCEAuthoritiesMotifs, MEDIAN$NormCEAuthoritiesMotifs, SD$NormCEAuthoritiesMotifs) colnames(Nresultsmotifs_summary) <- c(\"DATE\",\"N\",\"MEAN\",\"MEDIAN\",\"SD\") Nresultsmotifs_summary$SE <- Nresultsmotifs_summary$SD / sqrt(Nresultsmotifs_summary$N) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot mean and median H(A|M)/H(A) per motif per period [CODESPLIT] require(ggplot2) ggmean <- ggplot(Nresultsmotifs_summary,aes(x=DATE,y=MEAN)) + labs(title =  \"P1: Mean normalized conditional entropy of authorities given motifs across motifs\", x = \"Year BCE\", y = \"mean H(A|M)/H(A) across motifs\") + scale_x_reverse() + geom_errorbar(aes(ymin=Nresultsmotifs_summary$MEAN-Nresultsmotifs_summary$SE, ymax=Nresultsmotifs_summary$MEAN+Nresultsmotifs_summary$SE),width=.1) + geom_line() + geom_point() ggmean ggmedian <- ggplot(Nresultsmotifs_summary,aes(x=DATE,y=MEDIAN)) + labs(title =  \"P1: Median normalized conditional entropy of authorities given motifs across motifs\", x = \"Year BCE\", y = \"median H(A|M)/H(A) across motifs\") + scale_x_reverse() + geom_line() + geom_point() ggmedian ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "In renamed \"gender\" variable replace \"son\" with \"Male\" etc.  [CODESPLIT] StataGDIM$gender <- gsub('son', 'Male', StataGDIM$gender) StataGDIM$gender <- gsub('daughter', 'Female', StataGDIM$gender) StataGDIM$gender <- gsub('all', 'Mixed', StataGDIM$gender) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create cohort (decade of birth) variable in ACE data [CODESPLIT] StataACE <- StataACE %>% mutate(cohort = as.integer(dob_midrange/10)*10)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create short cohort ID by concatenating two character vectors. cohortIDshort is created to reduce clutter in multipanel figures [CODESPLIT] ETmerged$cohort_short <- substring(ETmerged$cohort, 3) ETmerged$cohortIDshort <- paste(ETmerged$wbcode, ETmerged$cohort_short, sep = \"\", collapse = NULL) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Inverse variance weightings for variance of h2, c2 and e2  [CODESPLIT] ETmerged<-ETmerged %>% mutate (h2weight = 1/h2var) ETmerged<-ETmerged %>% mutate (c2weight = 1/c2var) ETmerged<-ETmerged %>% mutate (e2weight = 1/e2var)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Manually derive pvalues from tvalues for Alt_Norway (for a twosided ttest): [CODESPLIT] 2*pt(abs(-1.0320512), df=14,lower.tail=FALSE) 2*pt(abs(2.48367345), df=14,lower.tail=FALSE) 2*pt(abs(-2.828329), df=14,lower.tail=FALSE) 2*pt(abs(1.2811099), df=14,lower.tail=FALSE) 2*pt(abs(3.4124371), df=14,lower.tail=FALSE) 2*pt(abs(0.6433726), df=14,lower.tail=FALSE) 2*pt(abs(-1.0321), df=14,lower.tail=FALSE) 2*pt(abs(2.4837), df=14,lower.tail=FALSE) 2*pt(abs(-2.8283), df=14,lower.tail=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlations separately for each partner partner 1 [CODESPLIT] chart.Correlation(data_dyadic[ , c(\"pes_1\",\"p_self_1\",\"d_self_1\",\"p_other_1\",\"d_other_1\")], use = \"pairwise.complete.obs\", pch = 20, histogram = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Remove\" AoI > 1200 or < 50 ms   [CODESPLIT] prevalues_gaze <- sum(PAST_O_19$gazedur > 0, na.rm=T)+sum(PAST_M_19$gazedur > 0, na.rm=T)+ sum(PRES_M_19$gazedur > 0, na.rm=T)+sum(PRES_O_19$gazedur > 0, na.rm=T) prevalues_fix <- sum(PAST_O_19$fixdur > 0, na.rm=T)+sum(PAST_M_19$fixdur > 0, na.rm=T)+ sum(PRES_M_19$fixdur > 0, na.rm=T)+sum(PRES_O_19$fixdur > 0, na.rm=T) PAST_M_19$gazedur[PAST_M_19$gazedur  > 1200 | PAST_M_19$gazedur < 50] <- NaN PAST_M_19$fixdur[PAST_M_19$fixdur > 1200 | PAST_M_19$fixdur < 50] <- NaN PRES_O_19$gazedur[PRES_O_19$gazedur > 1200 | PRES_O_19$gazedur < 50] <- NaN PRES_O_19$fixdur[PRES_O_19$fixdur > 1200 | PRES_O_19$fixdur < 50] <- NaN PRES_M_19$gazedur[PRES_M_19$gazedur > 1200 | PRES_M_19$gazedur < 50] <- NaN PRES_M_19$fixdur[PRES_M_19$fixdur > 1200 | PRES_M_19$fixdur < 50] <- NaN PAST_O_19$gazedur[PAST_O_19$gazedur > 1200 | PAST_O_19$gazedur < 50] <- NaN PAST_O_19$fixdur[PAST_O_19$fixdur > 1200 | PAST_O_19$fixdur < 50] <- NaN postvalues_gaze <- sum(PAST_O_19$gazedur > 0, na.rm=T)+sum(PAST_M_19$gazedur > 0, na.rm=T)+ sum(PRES_M_19$gazedur > 0, na.rm=T)+sum(PRES_O_19$gazedur > 0, na.rm=T) postvalues_fix <- sum(PAST_O_19$fixdur > 0, na.rm=T)+sum(PAST_M_19$fixdur > 0, na.rm=T)+ sum(PRES_M_19$fixdur > 0, na.rm=T)+sum(PRES_O_19$fixdur > 0, na.rm=T) dataloss_fix <- prevalues_fix-postvalues_fix dataloss_fix_percentage <- (1-(postvalues_fix/prevalues_fix))*100 dataloss_gaze <- prevalues_gaze-postvalues_gaze dataloss_gaze_percentage <- (1-(postvalues_gaze/prevalues_gaze))*100 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Distributional properties of the individual items for the Germanspeaking sample [CODESPLIT] jmv::descriptives( data = DataGerman, vars = vars(DMW1, DMW2, DMW3, DMW4, SMW1, SMW2, SMW3, SMW4, SBPS1, SBPS2, SBPS3, SBPS4, SBPS5, SBPS6, SBPS7, SBPS8), freq = TRUE, hist = TRUE, violin = TRUE, skew = TRUE, kurt = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Distributional properties of the individual items for the US sample [CODESPLIT] jmv::descriptives( data = DataUS, vars = vars(DMW1, DMW2, DMW3, DMW4, SMW1, SMW2, SMW3, SMW4, SBPS1, SBPS2, SBPS3, SBPS4, SBPS5, SBPS6, SBPS7, SBPS8), freq = TRUE, hist = TRUE, violin = TRUE, skew = TRUE, kurt = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Network analysis using EBICglasso [CODESPLIT] n1<-estimateNetwork(DataAll, default= \"EBICglasso\") plot(n1, groups = gr, nodeNames = names, legend.cex=.35) centrality_auto(n1, weighted = TRUE, signed = TRUE) centralityPlot(n1, include =c(\"Betweenness\",\"Closeness\", \"Strength\")) print(n1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Network based on correlations [CODESPLIT] n1a<-estimateNetwork(DataAll, default= \"cor\") plot(n1a, groups = gr, legend.cex=.35) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Network analysis using EBICglasso with the Germanspeaking sample including the openmindedness scale [CODESPLIT] gr2<- list('Deliberate MW'=c(1:4), 'Spontaneous MW'=c(5:8), 'Boredom'=c(9:16), 'Open-Mindedness'=c(17:22)) names2<-c(\"I allow my thoughts to wander on purpose\", \"I enjoy mind-wandering\", \"I find mind-wandering is a good way to cope with boredom\", \"I allow myself to get absorbed in pleasant fantasy\", \"I find my thoughts wandering spontaneously\", \"When I mind-wander my thoughts tend to be pulled from topic to topic\", \"It feels like I donÃ¢ÂÂt have control over when my mind wanders\", \"I mind wander even when IÃ¢ÂÂm supposed to be doing something else\", \"I often find myself at Ã¢ÂÂloose ends,Ã¢ÂÂ not knowing what to do\", \"I find it hard to entertain myself\", \"Many things I have to do are repetitive and monotonous\", \"It takes more stimulation to get me going than most people\", \"I donÃ¢ÂÂt feel motivated by most things that I do\", \"In most situations, it is hard for me to find something to do or see to keep me interested\", \"Much of the time, I just sit around doing nothing\", \"Unless I am doing something exciting, even dangerous, I feel half-dead and dull\", \"Has few artistic interests\", \"Is complex, a deep thinker\", \"Is original, comes up with new ideas\", \"Is fascinated by art, music, or literature\", \"Has little interest in abstract ideas\", \"Has little creativity\") n2<-estimateNetwork(DataGerman, default= \"EBICglasso\") plot(n2, groups = gr2, nodeNames = names2, legend.cex=.35) centrality_auto(n2, weighted = TRUE, signed = TRUE) centralityPlot(n2, include =c(\"Betweenness\",\"Closeness\", \"Strength\")) print(n2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Extract the vectors of analysed characteristics of Partner and Nonbiological father from the dataset. [CODESPLIT] text<-paste(\"trait_p<-data$Partner_\",char,sep=\"\") eval(parse(text=text)) text<-paste(\"trait_f<-data$Nonbiol_\",char,sep=\"\") eval(parse(text=text)) text<-paste(\"trait_pb<-dbiol$Partner_\",char,sep=\"\") eval(parse(text=text)) text<-paste(\"trait_b<-dbiol$Biol_\",char,sep=\"\") eval(parse(text=text)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Cohen's d First create the equivalent arrangement for the stimation of differences if the father is biologicla and present [CODESPLIT] muB<-rep(mean(diffb,na.rm=T),15) sdB<-rep(sd(diffb,na.rm=T),15) nB<-rep(length(diffb[!is.na(diffb)]),15) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plots also when there is no interaction and also squared terms (potentially involved in interactions) create data frame storing the observed mean values per combination of the values of the grided covariate and factor levels in the data create vectors denoting the grid cell center values for the two covariate: [CODESPLIT] if(!is.na(grid.resol)){ xvar=seq(min(plot.data[,covariate]), max(plot.data[,covariate]), length.out=grid.resol) bin.x=cut(x=plot.data[,covariate], breaks=xvar, include.lowest=T, labels=F) bin.x=min(xvar)+diff(xvar[1:2])/2+(bin.x-1)*diff(xvar[1:2]) }else{ bin.x=plot.data[, covariate] } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove rows where wkl_uuid is NA: [CODESPLIT] VS <- VS[!is.na(VS$wkl_uuid), ] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate data assuming an underlying VAR(1) process      Function to generate data assuming an underlying VAR(1) process  the input is N the number of samples, T.days and T.beeps are the number of days and the number of beeps ESM is conducted (Total number of assessments is T.days x T.beeps)  Psi is a matrix with the fixed regression weights, mu is a vector with the fixed intercepts, var.Psi is variance of the random regression weights (i.e., assumed to be equal),  For each participant, it is checked whether the matrix Psi conforms the assumption of stationary time series  with the absolute value of the maximum eigenvalue smaller than 1  The distribution of the random effects resembles a truncated multivariate normal distribution.  This function simulates data from a VAR(1) model  [CODESPLIT] Data.VAR.Fixed = function(N,T,Psi,cor.Sigma){ p = ncol(Psi) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " remove layers with gtype_class \"MF\" (only three distinct MF layers in data set!)  [CODESPLIT] toDiscard <- which(CM$gtype_class == \"MF\") if (length(toDiscard) > 0) CM <- CM[-toDiscard, ] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Testing participant heterogeneity  chisquare test: [CODESPLIT] testHetChi(freq = \"data/data_retrieval.csv\", tree = c(\"E\",\"E\",\"E\", \"U\",\"U\",\"U\", \"N\",\"N\",\"N\") ) testHetChi(freq = \"data/data_encoding.csv\", tree = c(\"E\",\"E\",\"E\", \"U\",\"U\",\"U\", \"N\",\"N\",\"N\") ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fitting a betaMPT model [CODESPLIT] m.retrieval.beta <- betaMPT(eqnfile=\"model/2htsm.eqn\", data = \"data/data_retrieval.csv\", restrictions = \"model/restrictions.txt\", modelfilename = \"results/2htsm_betaMPT.jags\", transformedParameters = list(\"deltaDd=D1-d1\"), parEstFile = \"results/results_retrieval_betaMPT.txt\", n.chain = 4, n.iter = 50000, n.adapt = 10000, n.burnin = 10000, n.thin = 10, ppp = 5000, dic = TRUE ) summary(m.retrieval.beta) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  Withinsubject tests  Example: twohigh threshold model (2HTM, included in TreeBUGS)  [CODESPLIT] htm <- system.file(\"MPTmodels/2htm.eqn\", package=\"TreeBUGS\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot ranked correlations [CODESPLIT] cor_dat <- corr_cross(ogdatasub, max_pvalue = 0.05, # display only significant correlations (at 5% level) pvalue =T, plot = F, top = 10 # display top 10 correlations (by correlation coefficient) ) |> arrange(abs(corr)) |> mutate(lab = paste(key, mix, sep = \" + \"), lab = factor(lab)) cor_dat |> ggplot(aes(y = lab, x = abs(corr), fill = as.factor(sign(corr)))) + geom_col() + scale_fill_manual(values = c(\"#E64B3D\", \"#0092B2\")) + labs(title=\"\", subtitle = \"\", caption=\"Note: Red bars indicate a negative relationship\") + geom_text(aes(label = signif(.data$corr, 2)), colour = \"#FFFFFF\", size = 3.5, hjust = 1.1) + scale_x_continuous(limits=c(0, .4), position=\"top\") + scale_y_discrete(labels=c(\"fst + cst\" = \"Difficult Financial Situation & Cost of Food\", \"rsp + spp\" = \"Loved Ones' Needs & Lack of Support\", \"pdd + ppd\" = \"Positive Experiences with Veg*nism & Positive Elements Post-Diet\", \"eft + prc\" = \"Meal Complexity & Need for Specialized Equipment or Planning\", \"fds + oth\" = \"Discontent and Cravings for Animal Products & Other Obstacles\", \"rsp + eft\" = \"Loved Ones' Needs & Meal Complexity\", \"fph + cst\" = \"Challenging Purchase Requirements & Cost of Food\", \"avl + fph\" = \"Limited Access to Veg*n Options & Challenging Purchase Requirements\", \"med + con\" = \"Professional Medical Advice & Difficulty Managing Health\", \"con + com\" = \"Difficulty Managing Health & Commitment Difficulties\", \"fcn + prc\" = \"Lack of Veg*n Knowledge & Need for Specialized Equipment or Planning\")) + theme_minimal() + theme(axis.title = element_blank(), legend.position = \"none\") cor_dat ggsave(\"Correlations.png\", width=8, height=6) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Predictive performance plot   ( model size selection plot) [CODESPLIT] stat_pretty <- setNames(nm = proj_evalstats) stat_pretty <- toupper(stat_pretty) stopifnot(identical(proj_evalstats, \"mlpd\")) ggeval <- plot(C_cvvs, stats = proj_evalstats, deltas = TRUE, ranking_nterms_max = NA) ggeval <- ggeval + facet_null() ggeval <- ggeval + scale_y_continuous( sec.axis = sec_axis(~ exp(.), name = bquote(Lambda*\" \"*\"GMPD\")) ) ggeval <- ggeval + labs(y = bquote(Delta*\" \"*.(stat_pretty))) print(ggeval) ggsave(file.path(\"output\", out_folder, paste0(plot_prefix, \"projpred_search_deltas.jpeg\")), width = 7, height = 7 * 0.618) saveRDS(last_plot(), file = file.path(\"output\", out_folder, paste0(plot_prefix, \"projpred_search_deltas.rds\"))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "save attribute level for option i in trial row.id [CODESPLIT] wide.data[[paste0(\"opt\",i,att.names[k])]][row.id]=feat.ij ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "save position of attribute j [CODESPLIT] wide.data[[paste0(\"pos\",att.names[k])]][row.id]=j ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "save attribute of position k [CODESPLIT] wide.data[[paste0(\"pos\",j)]][row.id]=att.names[k] } } } if(y$selected[i]==1) wide.data$response[row.id]=i } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1. recode RT for trials with slow (missed) response as NA [CODESPLIT] data$rt[data$trialError == \" Slow\"] <- NA ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ttests to compare composite 'animal protection behaviors' in people who have vs. have not experienced each (16 types total) [CODESPLIT] t.test(beh_animalprotect_comp ~ graphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp~ nongraphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ person_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ leaflet_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ news_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ social_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ humaneed_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ documentary_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ book_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ celebrity_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ ad_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ challenge_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ labels_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp~ labeled_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ ndisprotest_exp_buc, data=data, var.equal = TRUE) t.test(beh_animalprotect_comp ~ disprotest_exp_buc, data=data, var.equal = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "FDR correction for ttests that compared composite 'animal protection behaviors' between people who have and have not experienced each advocacy type [CODESPLIT] adjusted_pvalues_protectionbehaviors <- data %>% summarise(graphic_exp_buc = t.test(beh_animalprotect_comp ~ graphic_exp_buc, var.equal = TRUE)$p.value, nongraphic_exp_buc = t.test(beh_animalprotect_comp ~ nongraphic_exp_buc, var.equal = TRUE)$p.value, person_exp_buc = t.test(beh_animalprotect_comp ~ person_exp_buc, var.equal = TRUE)$p.value, leaflet_exp_buc = t.test(beh_animalprotect_comp ~ leaflet_exp_buc, var.equal = TRUE)$p.value, news_exp_buc = t.test(beh_animalprotect_comp ~ news_exp_buc, var.equal = TRUE)$p.value, social_exp_buc = t.test(beh_animalprotect_comp ~ social_exp_buc, var.equal = TRUE)$p.value, humaneed_exp_buc = t.test(beh_animalprotect_comp ~ humaneed_exp_buc, var.equal = TRUE)$p.value, documentary_exp_buc = t.test(beh_animalprotect_comp ~ documentary_exp_buc, var.equal = TRUE)$p.value, book_exp_buc = t.test(beh_animalprotect_comp ~ book_exp_buc, var.equal = TRUE)$p.value, celebrity_exp_buc = t.test(beh_animalprotect_comp ~ celebrity_exp_buc, var.equal = TRUE)$p.value, ad_exp_buc = t.test(beh_animalprotect_comp ~ ad_exp_buc, var.equal = TRUE)$p.value, challenge_exp_buc = t.test(beh_animalprotect_comp ~ challenge_exp_buc, var.equal = TRUE)$p.value, labels_exp_buc = t.test(beh_animalprotect_comp ~ labels_exp_buc, var.equal = TRUE)$p.value, labeled_exp_buc = t.test(beh_animalprotect_comp ~ labeled_exp_buc, var.equal = TRUE)$p.value, ndisprotest_exp_buc = t.test(beh_animalprotect_comp ~ ndisprotest_exp_buc, var.equal = TRUE)$p.value, disprotest_exp_buc = t.test(beh_animalprotect_comp ~ disprotest_exp_buc, var.equal = TRUE)$p.value) %>% gather(\"Advocacy\",\"p_value\") %>% mutate(p_fdr = p.adjust(p_value, method = \"fdr\", n = length(p_value))) %>% print() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ttests to compare composite 'animal consumer behaviors' in people who have vs. have not experienced each (16 types total) [CODESPLIT] t.test(beh_consumer_comp ~ graphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp~ nongraphic_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ person_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ leaflet_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ news_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ social_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ humaneed_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ documentary_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ book_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ celebrity_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ ad_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ challenge_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ labels_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ labeled_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ ndisprotest_exp_buc, data=data, var.equal = TRUE) t.test(beh_consumer_comp ~ disprotest_exp_buc, data=data, var.equal = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "prepare animal protection behaviors for stacked graph [CODESPLIT] mean_CI_protect <- data %>% group_by(advocacytype, experienced) %>% summarize(n = n(), protect_mean = mean(beh_animalprotect_comp), protect_lci = t.test(beh_animalprotect_comp, conf.level = 0.95)$conf.int[1], protect_uci = t.test(beh_animalprotect_comp, conf.level = 0.95)$conf.int[2]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "add asterisks to advocacy types that had sig different scores between experienced vs. not experienced from ttests [CODESPLIT] mean_CI_protect <- mean_CI_protect %>% mutate(advocacytype = case_when(advocacytype == \"ad\" ~ \"Anuncio o Valla Publicitaria*\", advocacytype == \"book\"  ~ \"Libro*\", advocacytype == \"celebrity\"  ~ \"Celebridad*\", advocacytype == \"challenge\"  ~ \"Reto de Evitar la Carne*\", advocacytype == \"disprotest\" ~ \"Protesta Disruptiva*\", advocacytype == \"graphic\"  ~ \"Video GrÃ¡fico*\", advocacytype == \"humaneed\"  ~ \"EducaciÃ³n en el Aula de Clases*\", advocacytype == \"labeled\"  ~ \"InformaciÃ³n Educativa Sobre las Etiquetas de Bienestar*\", advocacytype == \"labels\"  ~ \"Etiquetas Vegano/De Base Vegetal*\", advocacytype == \"ndisprotest\"  ~ \"Protesta No Disruptiva*\", advocacytype == \"news\" ~ \"ArtÃ­culo de Noticias*\", advocacytype == \"nongraphic\" ~ \"Video No GrÃ¡fico*\", advocacytype == \"person\" ~ \"DivulgaciÃ³n Boca a Boca *\", advocacytype == \"social\" ~\"PublicaciÃ³n en Redes Sociales o en un Blog*\", advocacytype == \"documentary\"  ~ \"Documental*\", advocacytype == \"leaflet\" ~ \"Folleto o Volante*\")) mean_CI_consume <- mean_CI_consume %>% mutate(advocacytype = case_when(advocacytype == \"ad\" ~ \"Anuncio o Valla Publicitaria*\", advocacytype == \"book\"  ~ \"Libro*\", advocacytype == \"celebrity\"  ~ \"Celebridad*\", advocacytype == \"challenge\"  ~ \"Reto de Evitar la Carne*\", advocacytype == \"disprotest\" ~ \"Protesta Disruptiva*\", advocacytype == \"graphic\"  ~ \"Video GrÃ¡fico*\", advocacytype == \"humaneed\"  ~ \"EducaciÃ³n en el Aula de Clases*\", advocacytype == \"labeled\"  ~ \"InformaciÃ³n Educativa Sobre las Etiquetas de Bienestar*\", advocacytype == \"labels\"  ~ \"Etiquetas Vegano/De Base Vegetal*\", advocacytype == \"ndisprotest\"  ~ \"Protesta No Disruptiva*\", advocacytype == \"news\" ~ \"ArtÃ­culo de Noticias*\", advocacytype == \"nongraphic\" ~ \"Video No GrÃ¡fico*\", advocacytype == \"person\" ~ \"DivulgaciÃ³n Boca a Boca *\", advocacytype == \"social\" ~\"PublicaciÃ³n en Redes Sociales o en un Blog*\", advocacytype == \"documentary\"  ~ \"Documental*\", advocacytype == \"leaflet\" ~ \"Folleto o Volante*\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "factorize advocacy type for graph [CODESPLIT] mean_CI_protect <- mean_CI_protect %>% mutate(advocacytype = factor(advocacytype, bar_order)) mean_CI_consume <- mean_CI_consume %>% mutate(advocacytype = factor(advocacytype, bar_order)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "grouped bar graph for animal protection behaviors [CODESPLIT] ggplot(mean_CI_protect, aes(x= advocacytype, fill=experienced, y=protect_mean)) + geom_col(width = 0.8, position=\"dodge\") + coord_flip(ylim=c(1,5)) + geom_errorbar(aes(x= advocacytype, ymin = protect_lci, ymax = protect_uci), width=0.4, colour = \"black\", position=position_dodge(.8)) + geom_text(aes(label = format(round(protect_mean, 1)), y = protect_uci), hjust = -0.2,# nudge_y = 2, size = 3, position = position_dodge(width = 1)) + labs(y= \"PuntuaciÃ³n Media del Comportamiento de ProtecciÃ³n de los Animales\", x = \"Tipo de Defensa\", caption = \"Un asterisco (*) indica que hubo una diferencia estadÃ­sticamente significativa entre los grupos (todos los ps < 0,05) despuÃ©s de haber corregido mediante FDR. Para mÃ¡s detalles sobre cÃ³mo se llevaron a cabo estos anÃ¡lisis, vÃ©ase el Material Complementario.\") + theme(legend.title = element_blank(), legend.position = \"bottom\", panel.background = element_rect(\"white\"), panel.border = element_rect(fill = NA), panel.grid.major.x = element_line(\"grey\")) + scale_fill_manual(values = c(\"Experimentado\" = \"#c47020\", \"No Experimentado\" = \"#F68D29\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We create a list of parameters to visualize and fill the labels etc to the .txt file in any table editor (Excel) [CODESPLIT] write.table(data.frame(nam=rep.par2),\"params_empty.txt\",row.names=F,sep=\"\\t\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot contrasts estimates [CODESPLIT] plot(NULL,ylim=c(max(dec$y)+const1+const2,min(dec$y)-const1),type=\"n\",xaxs=\"i\",yaxs=\"i\",xaxt=\"n\",yaxt=\"n\",xlim=xlims[[block]],bty=\"n\") abline(h=dec$y,col=col.grid,lty=1,lwd=lwd.grid) segments(axes[[block]],max(dec$y)+const1,axes[[block]],max(dec$y)+tic+const1,lwd=lwd.ax,col=col.ax) text(axes[[block]],max(dec$y)+tic+const1+ofs,labels=axes[[block]],col=col.ax,cex=0.9,font=2) lines(range(axes[[block]]),rep(max(dec$y)+const1,2),lwd=lwd.ax,col=col.ax) segments(axes[[block]],min(dec$y)-const1,axes[[block]],max(dec$y)+const1,lwd=lwd.v,col=col.ax,lty=3) segments(0,min(dec$y)-const1,0,max(dec$y)+const1,lwd=lwd.v,col=col.ax,lty=1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Draw density polygons density areas are scaled within each block. [CODESPLIT] area<-0.25*diff(xlims[[block]]) for(i in 1:nrow(dec)){ thispost<-postsc[,dec$n[i]] dens<-density(thispost) polX<-c(dens$x,rev(dens$x)) polY<-c(dens$y,rev(-dens$y)) ar1<-abs(polyarea(polX,polY)) perc<-area/ar1 polygon(polX,polY*perc+dei$y[i],col=dec$hex[i],border=col.pol,lwd=lwd.pol) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlations between csd_mean and averaged csds for Big Five traits   (mentioned in the text in Appendix A): [CODESPLIT] psych::corr.test( df2$n_csd,df2$csd_mean_n ) psych::corr.test( df2$e_csd,df2$csd_mean_e ) psych::corr.test( df2$o_csd,df2$csd_mean_o ) psych::corr.test( df2$a_csd,df2$csd_mean_a ) psych::corr.test( df2$c_csd,df2$csd_mean_c ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlations between variability and wellbeing measures: [CODESPLIT] names_var <- c(\"csd_mean_n\",\"csd_mean_e\",\"csd_mean_o\",\"csd_mean_a\",\"csd_mean_c\", \"sccs\",\"csd_nob\",\"simpson_csd\",\"rses\",\"swls\",\"pa\",\"na\") psych::corr.test( df2[,names_var] ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " catch situation where Sx is missing  so we cannot assess response to treatment  [CODESPLIT] if ( nrow( thisSx ) < 1 ) { thisRow$status.sx <- NA noFurtherFlag <- 1    ## tell rest of loop to not bother thisRow$missing.Sx <- 1 }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Descriptive analysis: counting occurrence of bodybased  units and coded themes.   Step 1. Analyse frequencies of bodybased units of measure  Ensure data is in character format [CODESPLIT] df$`Body dimension` <- as.character(df$`Body dimension`)", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Repeating confirmatory analyses with extra exclusions: in addition to excluding participants who did not complete the study, and excluding timeuntilguess and arithmeticsolvingtimes more than 5 standard deviations away from the mean, this analysis also excludes participants who removed 0 tiles when guessing.  Exclude participants who did not complete study  [CODESPLIT] d.conf_1_2.complete <- d.conf_1_2.b[complete.cases(d.conf_1_2.b),] d.conf_3_math.complete <- d.conf_3_math.b[complete.cases(d.conf_3_math.b),] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove rows with timeuntilguess values that are more than 5 standard deviations away from the mean [CODESPLIT] agg <- aggregate(ElapsedTime_Guess ~ Guess_Number + ID_Player + Effort + Competition, data=d.conf_1_2.complete, FUN = mean) sd5.times <- mean(agg$ElapsedTime_Guess) + (5 * sd(agg$ElapsedTime_Guess)) nrow(agg[agg$ElapsedTime_Guess > sd5.times,]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove rows with arithmeticsolvingtimes that are more than 5 standard deviations away from the mean [CODESPLIT] sd5.times <- mean(d.conf_3_math.complete$ElapsedTime_Math) + (5 * sd(d.conf_3_math.complete$ElapsedTime_Math)) nrow(d.conf_3_math.complete[d.conf_3_math.complete$ElapsedTime_Math > sd5.times,]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exclude participants who revealed 0 tiles [CODESPLIT] d.conf_1_2.complete_no0 <- d.conf_1_2.complete[d.conf_1_2.complete$TilesRevealed > 0,] d.conf_3_math.complete_no0 <- d.conf_3_math.complete[d.conf_3_math.complete$TilesRevealed > 0,] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compare simpler models with sex and sex*effort [CODESPLIT] df <- compare(m_accuracy_guess_nointer, m_accuracy_componly, m_accuracy_sex, m_accuracy_sex_effort_inter, m_accuracy_sex_effort_inter_CEinter) df <- round(df@output, 2) df %>% kable(caption = \"ACCURACY\")  %>% kable_styling(bootstrap_options = c(\"striped\", \"condensed\", \"responsive\"), full_width = TRUE) %>% column_spec(1:5, width = \"2cm\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "aggregate data for time elapsed per guess, by first removing duplicate times, then summing up total times per player  [CODESPLIT] d.reward.agg <- aggregate(ElapsedTime_Guess ~ Effort + Competition + Sex + Guess_Number + ID_Player, data = d.reward, FUN = mean) d.reward.agg <- aggregate(ElapsedTime_Guess ~ Effort + Competition + Sex + ID_Player, data = d.reward.agg, FUN = sum) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Bayes Factors  load additional libraries subset data into separate objects for frequentist analyses [CODESPLIT] d.math.agg.f <- d.math.agg d.math.agg.f$Competition <- as.factor(d.math.agg.f$Competition) d.math.agg.f$ID_Player <- as.factor(d.math.agg.f$ID_Player) d.math.agg.f$Sex <- as.factor(d.math.agg.f$Sex) d.conf.agg.f <- d.conf.agg d.conf.agg.f$Competition <- as.factor(d.conf.agg.f$Competition) d.conf.agg.f$Effort <- as.factor(d.conf.agg.f$Effort) d.conf.agg.f$ID_Player <- as.factor(d.conf.agg.f$ID_Player) d.conf.agg.f$Sex <- as.factor(d.conf.agg.f$Sex)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot AUC [CODESPLIT] ggplot(plot.data, aes(x = Target, y = AUC, color = Algorithm, fill = Algorithm, shape = Algorithm)) + geom_boxplot(width = 0.3,lwd = 1, aes(color = Algorithm, fill = Algorithm), alpha = 0.3, outlier.shape=NA, position=position_dodge(0.5)) + geom_point(position=position_jitterdodge(jitter.width = 0.1, dodge.width = 0.5), size = 0.5) + coord_flip() + geom_hline(yintercept = 0.5, color=\"black\", linetype = \"dashed\", size = 1) + theme_classic() + theme(text = element_text(size = 18), axis.title.y = element_blank(), legend.position = c(0.85, 0.5)) + scale_color_manual(values = c(\"#440154\", \"#5ec962\")) + scale_fill_manual(values = c(\"#440154\", \"#5ec962\")) + ylim(0.25, 1) + scale_x_discrete(labels=rev(c(\"Sociality\" = expression(paste(\"* \", bold(\"S\"), \"ociality\")), \"Deception\" = expression(paste(bold(\"D\"), \"eception\")), \"Negativity\" = expression(paste(bold(\"N\"), \"egativity\")), \"Positivity\" = expression(paste(\"* p\", bold(\"O\"), \"sitivity\")), \"Mating\" = expression(paste(\"* \", bold(\"M\"), \"ating\")), \"Adversity\" = expression(paste(bold(\"A\"), \"dversity\")), \"Intellectuality\" = expression(paste(\"* \", bold(\"I\"), \"ntellect\")), \"Duty\" = expression(paste(\"* \", bold(\"D\"), \"uty\"))))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Step 4*: Visualization of the ROC curves for \"core benchmark\" for the Online Supplemental Material  Duty  [CODESPLIT] p_Duty = autoplot(bmr$clone(deep = TRUE)$filter(task_ids = \"Duty\", learner_ids = c(\"lasso_Duty\", \"rf_Duty\")), type = \"roc\") + theme_classic() + theme(title = element_blank(), legend.position = \"none\", axis.title = element_text(size = 18), axis.text = element_text(size = 16)) + labs(y = \"Sensitivity\", x = \"1 - Specifity\") + geom_abline(intercept = 0, slope = 1, color = \"black\", linetype = \"dotted\", size = 0.8) + annotate(geom = \"text\", x = 0, y = 1, label = \"Duty\", hjust = 0, vjust = 1, size = 10) png(\"Figures_Tables/Figures/ROC_Curves/Duty.png\", width = 450, height = 400) p_Duty dev.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot ROC curves and caculate AUC (if desired) [CODESPLIT] plotIt <- FALSE currAUC <- plotAUC(allPs,allBFs,dt,numStudies,plotIt)  #currently does not save the AUCs ai <- ai+1 aucs[ai,1] <- sampleSizes[1] aucs[ai,2] <- effectSizes[2] aucs[ai,3] <- i #runNumber aucs[ai,4] <- currAUC[1] aucs[ai,5] <- currAUC[2] plotIt <- FALSE currAUC <- plotAUC(allPs2,allBFs2,dt,numStudies,plotIt)  #currently does not save the AUCs aucs$aucPhacked[ai] <- currAUC[1] aucs$aucBFhacked[ai] <- currAUC[2] if (currAUC[1] != currAUC[2]) { print(paste(effectSizes,i)) print(currAUC) } } #end for i runTimes } #end for ssa allDt <- allDt[-1,]  #get rid of initial dummy row that was just used to init allDt ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Stacked plots by SDT outcome   Plot by whether hacked or not plot hit, FA, miss, corrRejection for each effect size/sample size combo  [CODESPLIT] ap <- aggregate(hits ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) ap2 <- aggregate(fa ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(ap,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\",\"withHack\")) ap2 <- aggregate(miss ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(hFA,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\",\"withHack\")) ap2 <- aggregate(corrRej ~ criterion+sampleSize+effectSizes+withHack, data=allDt,mean) hFA <- merge(hFA,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\",\"withHack\")) hFA$hits <- hFA$hits / numStudies hFA$fa <- hFA$fa / numStudies hFA$corrRej <- hFA$corrRej / numStudies hFA$miss <- hFA$miss / numStudies ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot BF vs pvalues [CODESPLIT] plot(log(allPs),log(allBFs),bty=\"l\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create matrix of input variables with populationlevel effects: [CODESPLIT] C_datMM <- model.matrix( as.formula(paste(\"~\", paste(vpreds_noInt, collapse = \" + \"))), data = C_dat ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Set the latent regression coefficients for booklet 1, 5, 7 and 10 for reading and  booklet 4, 6, 9 and 11 for science to 0  [CODESPLIT] betas <- data.frame(var = c(which(names(con_dat) %in% paste0(\"bookid.\", c(1, 5, 7, 10))), which(names(con_dat) %in% paste0(\"bookid.\", c(4, 6, 9, 11)))), dim = rep(2:3, each = 4), value = 0) betas <- data.frame(var = c(which(names(con_dat) %in% paste0(\"bookid.\", c(1, 5, 7, 10))), which(names(con_dat) %in% paste0(\"bookid.\", c(4, 6, 9, 11)))), dim = rep(2:3, each = 4), value = 0) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute latent regression of the latent ability on the conditioning  variables (excl. the first column which is the student ID) [CODESPLIT] latreg <- tam.latreg(likeli, Y = con_dat[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = \"Ramsay\"), beta.fixed = as.matrix(betas)) } else { latreg <- tam.latreg(likeli, Y = con_dat[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = \"Ramsay\")) } latreg <- tam.latreg(likeli, Y = con_dat1[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = \"Ramsay\"), beta.fixed = as.matrix(betas)) } else { latreg <- tam.latreg(likeli, Y = con_dat1[, -1], pid = pid.sele, control = list(maxiter = iter.2, acceleration = \"Ramsay\")) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Draw 5 plausible values for each student out of the resulting distribution  which is assumed to be normal distributed [CODESPLIT] pvs    <- tam.pv(latreg, nplausible = 5, normal.approx = T, samp.regr = samp.regr.opt) } else { pvs    <- tam.pv(latreg_md, nplausible = 5, normal.approx = T, samp.regr = samp.regr.opt) } else { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Extract the regression coefficients of the conditioning variables, because  they are fixed for the core domains in the next step at that value  [CODESPLIT] reg.coefs <- cbind(rep(1:dim(latreg$beta)[1], 3), rep(1:3, each = dim(latreg$beta)[1]), c(latreg$beta[, 1], latreg$beta[, 2], latreg$beta[, 3])) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Extract IRT likelihood of the second model (math, read and scie plus digital  domains)  [CODESPLIT] likeli_md <- IRT.likelihood(mod2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute latent regression of the latent ability on the conditioning  variables (excl. the first column which is the student ID). But this time  only the regression coefficients of the digital domains are computed freely.  The rest is fixed at the values of the first model [CODESPLIT] latreg_md <- tam.latreg(likeli_md, Y = con_dat[, -1],  pid = pid.sele, control = list(maxiter = iter.2, acceleration = \"Ramsay\"), beta.fixed = reg.coefs) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate the number of surveys completed per person [CODESPLIT] N <- as.data.frame(table(dat$id)) names(N) <- c(\"id\", \"N\") dat <- merge(dat, N, by = \"id\", all.x = T) table(dat$N, useNA = \"ifany\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " DOWNSAMPLING  Loop to downsample data based on minimum time and distance between points Time parameter is set in \"while(diff < time in minutes | dist > in meters)\" downsampled pts that were less than 60min apart unless frog moved more than 20m during the time window  [CODESPLIT] tracks_dsmpl <- data.frame() ids <- unique(tracks$id) for(i in ids) { traj = subset(tracks, tracks$id == i) for(i in 1:nrow(traj)) { diff <- difftime(traj$dt[i+1], traj$dt[i], units = \"mins\") delta_x <- traj$x_utm[i+1] - traj$x_utm[i] delta_y <- traj$y_utm[i+1] - traj$y_utm[i] dist <- sqrt(delta_x^2+delta_y^2) if(is.na(diff)) {break} while(diff <= 60 & dist < 20) { traj <- traj[-(i+1),] diff <- difftime(traj$dt[i+1], traj$dt[i], units = \"mins\") delta_x <- traj$x_utm[i+1] - traj$x_utm[i] delta_y <- traj$y_utm[i+1] - traj$y_utm[i] dist <- sqrt(delta_x^2+delta_y^2) if(is.na(diff)) {break} } } tracks_dsmpl <- bind_rows(tracks_dsmpl, traj) } tracks_dsmpl <- tracks_dsmpl %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = \"min\"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) tracks <- tracks_dsmpl ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Remove first row (tagging date) and last row (untagging day) per individual   Sort the daily, group by id and add index of id+date as a new variable  [CODESPLIT] daily %>% arrange(id) %>% group_by(id) %>% mutate(id_day = paste(id, date)) -> daily_grouped ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Join with summary stats [CODESPLIT] tracks_sum  <- left_join(tracks_sum, duration) tracks_sum  <- left_join(tracks_sum, relocs) tracks_sum  <- left_join(tracks_sum, days_tracked) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate average daily movement by behavioral category per individual for plotting [CODESPLIT] daily_beh_mean <- daily_beh %>% group_by(id, behavior) %>% dplyr::summarize(daily_dist = mean(daily_dist), max_dist = mean(max_dist), sex = first(sex), species = first(species)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "MCPs home range with 95% minimum convex ploygon (MCP95) [CODESPLIT] mcp95<- mcp(data_sp[\"id\"], percent = 95, unout = \"m2\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "95% contour with hpi pluging method Define countour level [CODESPLIT] cont=c(95)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Initiate empty df to store the metadarta [CODESPLIT] allfrogs_info <- data.frame(ID = numeric(0), Cont = numeric(0), frog_id = character(0)) allfrogs_info <- data.frame(ID = numeric(0), Cont = numeric(0), frog_id = character(0)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "BOXPLOTS: DAILY MOVEMENT  FIGURE EXPORT: daily movement by sex [CODESPLIT] daily_plot <-tracks_sum%>% ggplot(aes(x= sex, y=log(mean_cumul), fill = species)) + theme_bw(20) + geom_boxplot(width= 0.6, outlier.shape = NA) + geom_jitter(aes(group = sex), position=position_jitterdodge(0.4), shape = 21, stroke = 1, color = \"black\", size = 2, alpha = 0.6) + scale_color_manual(values=c(\"black\", \"black\")) + scale_fill_manual(values=c(\"#E7B800\", \"#0072B2\",\"#FC4E07\")) + theme(legend.position=\"none\") + labs(y = \"ln Daily movement (m)\") + facet_wrap(~species, labeller = labeller(species=c(\"afemo\" = \"A. femoralis\", \"dtinc\" = \"D. tinctorius\", \"osylv\" = \"O. sylvatica\"))) + theme(axis.title.x = element_blank(), axis.text.x = element_text(color = \"black\", size = 18), strip.text = element_text(face = \"italic\"), aspect.ratio = 4) + scale_x_discrete(labels= c(\"F\", \"M\"), expand = expansion(add = 1)) daily_plot ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Daily movement by sex and species, log transformed violin plots  [CODESPLIT] ggplot(daily_select, aes(x= sex, y=log(daily_dist))) + theme_bw(20) + geom_violin(aes(fill = species)) + scale_fill_manual(values=c(\"#E7B800\", \"#0072B2\",\"#FC4E07\")) + theme(legend.position=\"none\") + labs(y = \"ln Daily movement (m)\") + geom_jitter(aes(shape=sex, fill = species), position=position_jitterdodge(0.2), size = 4, alpha=0.2) + facet_wrap(~species, labeller = labeller( species=c(afemo =\"A. femoralis\", dtinc = \"D. tinctorius\", osylv = \"O. sylvatica\"))) + theme(axis.title.x = element_blank(), axis.text.x = element_text(color = \"black\", size = 18), strip.text = element_text(face = \"italic\")) + scale_x_discrete(labels= c(\"F\", \"M\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "convert attributes from character to numeric [CODESPLIT] for(i in 1:3) { data[[paste0(\"opt\",i,\"cost\")]] <- as.numeric(substring(data[[paste0(\"opt\",i,\"cost\")]], 3, 7)) data[[paste0(\"opt\",i,\"sides\")]] <- as.numeric(substring(data[[paste0(\"opt\",i,\"sides\")]], 1, 2)) data[[paste0(\"opt\",i,\"deliveryTime\")]] <- as.numeric(substring(data[[paste0(\"opt\",i,\"deliveryTime\")]], 2, 3)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Step 9: Create time difference variable, and the final dependent variable (media_success)  calculating the time difference between each documentpair  [CODESPLIT] result$date_diff <- as.Date(as.character(result$news_date), format=\"%Y-%m-%d\")- as.Date(as.character(result$date), format=\"%Y-%m-%d\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Remove trials with RT < 4710 ms (i.e., presses before disambiguating information in sentence)  [CODESPLIT] dataset_exp1 <- dataset_exp1[which(dataset_exp1$rt > 4710), ] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Add 'trackloss' column (if not looking at IA_1 or IA_2, then trackloss  1)  [CODESPLIT] dataset_exp1 <- within(dataset_exp1, { trackloss <- ifelse(average_target_sample_count_proportion == 0 & average_distractor_sample_count_proportion == 0, 1, 0) }) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Ttests (using participantaveraged data) [CODESPLIT] num_sub_exp1 <- length(unique((eyetrackingr.data.exp1$participant_number))) threshold_t_exp1 <- qt(p = 1 - .05 / 2, df = num_sub_exp1 - 1) # Pick threshold for t based on alpha = .05, two-tailed df_timeclust_exp1 <- make_time_cluster_data(response_time_exp1, test = \"t.test\", paired = TRUE, predictor_column = \"trial_condition_new\", threshold = threshold_t_exp1 ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "MSE for each participant [CODESPLIT] MSE.Sys.VAR.i = lapply(1:fold, function(k) MSE.k[[k]]$MSE.ki) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "use mutate and the dffit fit information to make the data for mal [CODESPLIT] mal_fit <- fit_num %>% mutate(fit = (s)/(1.196506 + s)) dex_fit <- fit_num %>% mutate(fit = (s)/(62.736799 + s)) lac_fit <- fit_num %>% mutate(fit = (s)/(43.839063 + s)) maldex_fit <- fit_num %>% mutate(fit = (s)/(7.253739 + s))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Logistic Mixed Effects Regression for Original Critical Targets [CODESPLIT] log_reg_data <- cards_long %>% filter(card == \"card_1\" | card == \"card_3\" | card == \"card_6\" | card == \"card_7\" | card == \"card_9\" | card == \"card_14\" ) %>% mutate( distance = case_when( card == \"card_1\" | card == \"card_3\" | card == \"card_6\"  ~ 0, card == \"card_7\" | card == \"card_9\" | card == \"card_14\" ~ 1 ), comf_acquaint = scale(comf_acquaint, scale = FALSE), comf_close = scale(comf_close, scale = FALSE), approp_acquaint = scale(approp_acquaint, scale = FALSE), approp_close = scale(approp_close, scale = FALSE) ) log_reg_data$amit_wording <- factor(log_reg_data$amit_wording, levels = c(\"original\", \"new\")) h1_melogreg <- glmer(choice ~ distance + amit_wording + (1 + distance|id) + (1|card), data = log_reg_data, family = binomial(link = \"logit\")) h1_melogreg_inter <- glmer(choice ~ distance * amit_wording + (1 + distance|id) + (1|card), data = log_reg_data, family = binomial(link = \"logit\")) h1_model_comparison <- anova(h1_melogreg, h1_melogreg_inter) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Highest density interval ' ' This is a function that will calculate the highest density interval from a ' posterior sample. ' ' The default is to calcualte the highest 95 percent interval. It can be used ' with any numeric vector instead of having to use one of the specific MCMC ' classes. This function has been adapted from John K. Kruschke (2011). Doing ' Bayesian Data Analaysis: A Tutorial with R and BUGS. ' ' @param x Numeric vector of a distribution of data, typically a posterior ' sample ' @param prob Width of the interval from some distribution. Defaults to `0.95`. ' @param warn Option to turn off multiple sample warning message Must be in the ' range of `[0,1]`. ' @return Numeric range ' @export ' @examples ' x < qnorm(seq(1e04, .9999, length.out1001)) ' hdi_95 < hdi(x, .95) ' hdi_50 < hdi(x, .50) ' ' hist(x, br50) ' abline(vhdi_95, col\"red\") ' abline(vhdi_50, col\"green\") ' ' x < exp(seq(pi * (1  (1/16)), pi, len  1000)) ' x < c(x, rev(x)[1]) ' x < c(x, x) ' plot(sort(x), type\"l\") ' plot(density(x, adjust0.25)) ' abline(vhdi(x, p.49), col2) ' abline(vhdi(x, p.50), col3)  [CODESPLIT] hdi <- function(x, prob=0.95, warn=TRUE) { if (anyNA(x)) { stop(\"HDI: \", \"x must not contain any NA values.\", call.=FALSE) } N <- length(x) if (N < 3) { if (warn) { warning(\"HDI: \", \"length of `x` < 3.\", \" Returning NAs\", call.=FALSE) } return(c(NA_integer_, NA_integer_)) } x_sort <- sort(x) window_size <- as.integer(floor(prob * length(x_sort))) if (window_size < 2) { if (warn) { warning(\"HDI: \", \"window_size < 2.\", \" `prob` is too small or x does not \", \"contain enough data points.\", \" Returning NAs.\", call.=FALSE) } return(c(NA_integer_, NA_integer_)) } lower <- seq_len(N - window_size) upper <- window_size + lower ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "test whether Condition predicts exclusion probability [CODESPLIT] Cond_exclude <- table(data$Cond, data$exclude) chisq.test(Cond_exclude) rm(Cond_exclude) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Fix the following script by:  Take our dataset (df) , and then  group it by our explanatory variable (Condition), and then  Summarise this dataset by  Creating a new variable in this summary dataset called Mean  Mean takes the mean of our Response variable, and ignores NA.  Repeat for standard deviation.  We've added on one more column in our summary dataset  N.  N is the number of observations per group. To do so, we want to add up  How many times our DV is not missing (per group).  When you think it is a complete, save the script. The name of the script  above will go from red to black. That is how you know you've saved it.  You can use ctrl+S to save, or click the floppy disk above.  We want to know how the means and standard deviations and Ns of Hire_Rating differed.  [CODESPLIT] sum_hire <-", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate R squared for each model [CODESPLIT] r.squaredGLMM(m1) #indegree r.squaredGLMM(m2) #outdegree r.squaredGLMM(m3) #betweenness r.squaredGLMM(m4) #outcloseness r.squaredGLMM(m5) #incloseness r.squaredGLMM(m6) #local clustering r.squaredGLMM(m7) #average shortest path length r.squaredGLMM(m8) #eigenvector r.squaredGLMM(m9) #outstrength r.squaredGLMM(m10) #instrength", "target": 1, "target_options": ["no_match", "match"]}
{"input": "split the data into two halves [CODESPLIT] d$ItemNo = parse_number(as.character(d$ItemNo)) de = d %>%  filter(ItemNo %% 2 == 0) do = d %>%  filter(ItemNo %% 2 == 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Zscore transform RTs, compute zRT difference, remove NAs for the 2 groups  [CODESPLIT] de <- de |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) do <- do |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) de <- de |> group_by(Subject) |> mutate (zRT = scale(Reaction.Time, center = T, scale = T), meanRT = mean(Reaction.Time)) |> select (1,2,4,5) |> pivot_wider(names_from = Condition, values_from = zRT) |> mutate (zrt.diff = R-P) |> filter(!is.na(zrt.diff)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit models for even and odd numbered items (Bayesian & Generalized LMM) [CODESPLIT] m_de = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = de, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) m_do = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = do, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) m_de = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = de, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) m_do = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = do, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit lmm to the whole dataset [CODESPLIT] m <- blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = d_trimmed, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) summary(m) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate shared and unique variance [CODESPLIT] x1 = rsq - rx2y^2 x2 = rsq - rx1y^2 x1_x2 = rsq - x1 - x2 c(x1, x2, x1_x2, rsq) } semipartial_corr = function(r12, r13, r23) { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create empty df [CODESPLIT] df <- data.frame( `rx1y` = numeric(), `rx2y` = numeric(), `rx1x2` = numeric(), `rx1y_x2` = numeric(), `rx2y_x1` = numeric(), `x1` = numeric(), `x2` = numeric(), `Common` = numeric(), `Total` = numeric() ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " > Left panel: QQplot (uniform distribution)  > Right panel: Residuals against predicted values;; shaded (due to sample size) with extreme residuals colored red, and   3) MAIN EFFECTS OF KEY VARIABLES   AspElevForm  [CODESPLIT] emmip(PrefRate, ~AspElevForm, type = \"response\", CIs = TRUE) (emm <- emmeans(PrefRate, specs = ~AspElevForm, type = \"response\")) pairs(emm) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Operator on x, rescale to be in y unit  [CODESPLIT] Hlist <- list() for(i in seq(nt)){ Hlist[[paste0(\"H\",i)]] <- with(data, t(apply(data[[i]]$sdfp, 1, function(x) x*data[[i]]$prior)) %*% A) } H<-bdiag(unlist(Hlist)) strl <- function(x){ x[x < max(x)/1000] <- 0 return(x) } H <- as(t(apply(H, 1, strl)), 'dgCMatrix')  #Make sparser ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Creates the variable 'distances', a 540 X 3 distance matrix between stimuli and  category prototypes.  [CODESPLIT] distances = as.matrix(pdist(coordinates, prototypes)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate similarity coefficient (mean of signif_line$similarity.scores) [CODESPLIT] mean(signif_line$similarity.scores[,2]) mean(coef_line$similarity.scores[,2]) mean(varimp_line$similarity.scores[,2]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate Priestley Taylor Ep [CODESPLIT] data$PT.Ep <- 1.26/lambda.w*deltav/(deltav+rho.air)*Rn", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot hit, FA, miss, corrRejection for each effect size/sample size combo  [CODESPLIT] ap <- aggregate(hits ~ criterion+sampleSize+effectSizes, data=allDt,mean) ap2 <- aggregate(fa ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(ap,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\")) ap2 <- aggregate(miss ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(hFA,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\")) ap2 <- aggregate(corrRej ~ criterion+sampleSize+effectSizes, data=allDt,mean) hFA <- merge(hFA,ap2,by=c(\"criterion\",\"sampleSize\",\"effectSizes\")) hFA$hits <- hFA$hits / numStudies hFA$fa <- hFA$fa / numStudies hFA$corrRej <- hFA$corrRej / numStudies hFA$miss <- hFA$miss / numStudies ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " plot errors only (FA and misses, separately)   Plot ROC dist   [CODESPLIT] aa <- aggregate(ROCdist ~ criterion, data=allDt,mean) ab <- aggregate(ROCdist ~ criterion, data=allDt,sd) ab$ci <- qnorm(.975) * ab$ROCdist / sqrt(numStudies) titleText <- ifelse(length(sSizesAll) < 2, paste(\"N=\",sSizesAll[1],\"d=\",effSzAll[1]), paste(\"firstES:\",effSzAll[1])) plot(seq(1:length(crits)),aa$ROCdist,bty=\"l\",xaxt=\"n\",xlab=\"Criterion for Statistical Significance\",ylab=\"Distance to Perfection\",pch=19,col=rainbow(length(crits)),cex=2,ylim=c(0,max(aa$ROCdist)+max(ab$ci)),main = titleText) axis(side=1,at=seq(1:length(crits)),labels = crits) for(i in 1:length(crits)) { segments(i,aa$ROCdist[i] - ab$ci[i],i,aa$ROCdist[i]+ab$ci[i]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate the variance and bias for the log(DR) [CODESPLIT] log.var <- ((res$se.tau2)^2)/4*1/(sum(w.star))^2*(sum(1/(res$vi + res$tau2)^2))^2 log.sd <- sqrt(log.var) bias <- 1/2*(res$se.tau2)^2*(1/2/sum(w.star)^2 - 1/sum(w.star)*sum(1/(res$vi + res$tau2)^3))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create the forest plot including extra information (study names, weights, observed effects and associated variances).  [CODESPLIT] f <- forest(res, col = \"blue\", border = \"blue\", ylim = c(-8,12), xlim = c(-4,5), pch = 19, slab = dat$study, showweights = FALSE, addfit = FALSE, refline = FALSE, ilab = cbind(paste(format(round(res$yi,2), nsmall = 2)) , paste(format(round(res$vi, 2), nsmall = 2)), paste(format(round(res$ni, 2))), paste(format(weights, nsmall = 2)), paste(format(weights.f, nsmall = 2))) , ilab.xpos = c(-3.15,-2.75, -2.35,-1.5,-0.75), ilab.pos = 2, efac = 0, digits = 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "STEP 2: Filter the data in overall.data table according to preregistered inclusion criteria    replace 0 with NA in outcome column [CODESPLIT] overall.data$LTScreenOut[overall.data$LTScreenOut == 0] <- NA #replace NAs with 0 for averaging in the next steps overall.data$LTObjectOut[overall.data$LTObjectOut == 0] <- NA #replace NAs with 0 for averaging in the next steps overall.data$FirstLookDurationObjectOut[overall.data$FirstLookDurationObjectOut == 0] <- NA #replace NAs with 0 for averaging in the next steps", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit indices function for the SEM and FA analyses [CODESPLIT] fa.CFI<-function(x){ nombre<-paste(x,\"CFI\",sep = \".\") nombre<- ((x$null.chisq-x$null.dof)-(x$STATISTIC-x$dof))/(x$null.chisq-x$null.dof) return(nombre) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check Descriptive statistics    gender breakdown [CODESPLIT] df <- eval_data %>% group_by(gender) %>% summarise(counts = n()) df", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Decompose variance  Estimate multilevel model (without predictors) [CODESPLIT] m_logistic <- lmer(estimate ~ 1 + (1|controls) + (1|age) + (1|year) + (1|age:year) + (1|age:controls) + (1|year:controls), data = results %>% rename(age = age_group)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Visualize four first principal components [CODESPLIT] fig.df <- data.frame(HR = c(FPCAdense$phi[,1], FPCAdense$phi[,2], FPCAdense$phi[,3], FPCAdense$phi[,4]), pc = c(rep(1, 51), rep(2, 51), rep(3, 51), rep(4, 51)), Time= rep(FPCAdense$workGrid, 4)) all.four <- ggplot(fig.df, aes(x=Time, y=HR))+geom_line()+ geom_vline(xintercept=.20, linetype=\"dashed\")+ facet_grid(.~pc)+ theme_bw() all.four ggsave(\"../figures/figure5.png\", width=6, height=4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "SPATIOTEMPORAL VARIABLES AND HOMING SUCCESS  Join territory center pts to the dataframe to calculate translocation distance and homing success [CODESPLIT] trajectory <- trajectory.df %>% left_join(territory_centers, by = \"id\")%>% mutate(trans_dist = sqrt((x_home - first(x_utm))^2+(y_home - first(y_utm))^2), trans_group = ifelse(trans_dist > 100, \"200m\", \"50m\"))%>% dplyr::select(id, sex, trans_group, trans_dist, dt, x_utm, y_utm, x_home, y_home, time_lag_min = time_diff, dist) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Tally the number relocations for each individual per day [CODESPLIT] trajectory.df %>% group_by(id, date) %>% count(id) -> relocs.perday ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot full circular info by sex  start with blank plot, then add groupspecific points  [CODESPLIT] dev.new();; par(mai = c(1, 1, 0.1,0.1)) par(mar=c(0.5, 0.5, 0.5, 0.5)) pdf(\"dt_circ_50m.pdf\") plot(corient50_m, bg = rgb(0, 0.749, 0.769), pch = 21, cex = 1.5, lwd = 2, stack = TRUE, bin = 60, xlim = c(-1.2, 1.2), ylim = c(-1.2, 1.2), sep = 0.05, shrink = 1, tcl.text = -0.125, control.circle=circle.control(lwd = 2)) ticks.circular(circular(seq(0,2*pi,pi/2)), tcl=0.075) par(new = T) plot(corient50_f, bg=rgb(0.973, 0.463, 0.427), pch = 21, cex=1.5, lwd = 2, stack=T, bins=60, sep = -0.05, shrink= 1.3, axes = FALSE, control.circle=circle.control(lwd = 1)) arrows.circular(mean(corient50_m), y = rho.circular(corient50_m), col = rgb(0, 0.749, 0.769), lwd = 5) arrows.circular(mean(corient50_f), y = rho.circular(corient50_f), col = rgb(0.973, 0.463, 0.427), lwd = 5) par(new = T) plot(corient50_m, col = NA, shrink= 2.5, axes = FALSE, control.circle=circle.control(lty = 2, lwd = 1)) ticks.circular(circular(seq(0,2*pi,pi/8)), tcl=0.2) lines(density.circular(corient50_m, bw=30), shrink= 1, col = rgb(0, 0.749, 0.769, 0.7), lwd=2, lty=1) lines(density.circular(corient50_f, bw=30), col = rgb(0.973, 0.463, 0.427, 0.7), lwd=2, lty=1) dev.new();; par(mai = c(1, 1, 0.1,0.1)) par(mar=c(0.5, 0.5, 0.5, 0.5)) pdf(\"dt_circ_200m.pdf\") plot(corient200_m, bg = rgb(0, 0.749, 0.769), pch = 21, cex = 1.5, lwd = 2, stack = TRUE, bin = 60, xlim = c(-1.2, 1.2), ylim = c(-1.2, 1.2), sep = 0.05, shrink = 1, tcl.text = -0.125, control.circle=circle.control(lwd = 2)) ticks.circular(circular(seq(0,2*pi,pi/2)), tcl=0.075) par(new = T) plot(corient200_f, bg=rgb(0.973, 0.463, 0.427), pch = 21, cex=1.5, lwd = 2, stack=T, bins=60, sep = -0.05, shrink= 1.3, axes = FALSE, control.circle=circle.control(lwd = 1)) arrows.circular(mean(corient200_m), y = rho.circular(corient200_m), col = rgb(0, 0.749, 0.769), lwd = 5) arrows.circular(mean(corient200_f), y = rho.circular(corient200_f), col = rgb(0.973, 0.463, 0.427), lwd = 5) par(new = T) plot(corient200_m, col = NA, shrink= 2.5, axes = FALSE, control.circle=circle.control(lty = 2, lwd = 1)) ticks.circular(circular(seq(0,2*pi,pi/8)), tcl=0.2) lines(density.circular(corient200_m, bw=30), shrink= 1, col = rgb(0, 0.749, 0.769, 0.7), lwd=2, lty=1) lines(density.circular(corient200_f, bw=30), col = rgb(0.973, 0.463, 0.427, 0.7), lwd=2, lty=1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Helper function to prepare raw keyboard data ' ' @author F. Bemmann ' @family Preprocessing function ' @description this function unfolds json data into one column per keyvalue pair ' @export [CODESPLIT] parseJsonColumnSensing = function(df, column_name){ parseJsonColumn = function(x){ str_c(\"[ \", str_c(x, collapse = \",\", sep=\" \"), \" ]\")  %>% jsonlite::fromJSON(flatten = T) %>% as_tibble() } df2 = df %>% select(user_uuid,client_event_id,!!column_name) %>% filter(!is.na(!!rlang::sym(column_name))) %>% map_dfc(.f = parseJsonColumn) %>% distinct() colnames(df2)[1:2] = c(\"user_uuid\", \"client_event_id\") df = left_join(df, df2, by = c(\"user_uuid\", \"client_event_id\")) df[,column_name] = NULL return(df) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Separate ANOVAs for each subscale  Pairwise ttests with adjusted alpha level of .05/4  .0125 (because there are four relevant comparisons) [CODESPLIT] summary(aov(value~Fan*time+Error(VP_t0/(time)),data=dat[dat$scale==\"Attentiveness\",])) pairwise.t.test(dat[dat$scale==\"Attentiveness\",]$value,paste(dat[dat$scale==\"Attentiveness\",]$Fan,dat[dat$scale==\"Attentiveness\",]$time),p.adj=\"none\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Models   .. Contrasts   (Reverse) Helmert for cond speech, cond mask  Treatment for speech_mask  Sum for group  Factor variables  [CODESPLIT] df <- df %>% mutate( group = factor(group), # oc, pd gender = factor(intake_gender), # f, m cond_mask = factor(cond_mask, levels = c(\"nm\",\"sm\",\"kn\")), cond_speech = factor(cond_speech, levels = c(\"habitual\",\"clear\",\"loud\")) ) %>% mutate(speech_mask = paste(cond_speech,cond_mask,sep=\"_\")) levels(df$group) # oc, pd levels(df$gender) # f, m levels(df$cond_speech) # habitual, clear, loud levels(df$cond_mask) # nm, sm, kn ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "4) Descriptives and data inspection   total duration on survey [CODESPLIT] stat.desc(working_file %>% group_by(pp) %>% slice(1) %>% pull(duration))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "bayesian ttests [CODESPLIT] ttestBF(wide$high, wide$neutral, paired = TRUE) ttestBF(wide$low, wide$neutral, paired = TRUE) ttestBF(wide$high, wide$low, paired = TRUE) formatC(3057348376, format = \"e\", digits = 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Confidence interval as a vector [CODESPLIT] result <- c(\"lower\" = vec_mean - error, \"upper\" = vec_mean + error) return(result) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Transform academic background and job category variables from character to factor [CODESPLIT] class(expertsample$acadgroup) expertsample$acadgroup<-as.factor(expertsample$acadgroup) class(expertsample$job_cat) expertsample$job_cat <- as.factor(expertsample$job_cat)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "graphical check for normal distribution of differences support (mean) [CODESPLIT] diff_support <- bothexpert$BECCS_support-bothexpert$DACCS_support hist(diff_support)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Wilcoxon tests support (mean) [CODESPLIT] wilcox.test(bothexpert$BECCS_support, bothexpert$DACCS_support, paired=TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlations separately for each condition [CODESPLIT] data_con <- data[data$Cond == \"Control\", ] data_ent <- data[data$Cond == \"Entitlement\", ] chart.Correlation(data_con[ , c(\"pes\",\"prestige\",\"dominance\",\"benign\",\"malicious\",\"pain\")], use = \"pairwise.complete.obs\", pch = 20, histogram = TRUE) chart.Correlation(data_ent[ , c(\"pes\",\"prestige\",\"dominance\",\"benign\",\"malicious\",\"pain\")], use = \"pairwise.complete.obs\", pch = 20, histogram = TRUE) rm(data_con,data_ent) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Estimate partial correlation matrices using fiml (partial cors and not glasso b/c different sample sizes leads to different glasso penalties) [CODESPLIT] net1hi <- cor1.edhi$cor %>% EBICglasso(., n = cor1.edhi$n) net2hi <- cor2.edhi$cor %>% EBICglasso(., n = cor2.edhi$n) net3hi <- cor3.edhi$cor %>% EBICglasso(., n = cor3.edhi$n) net4hi <- cor4.edhi$cor %>% EBICglasso(., n = cor4.edhi$n) net5hi <- cor5.edhi$cor %>% EBICglasso(., n = cor5.edhi$n) net6hi <- cor6.edhi$cor %>% EBICglasso(., n = cor6.edhi$n) net7hi <- cor7.edhi$cor %>% EBICglasso(., n = cor7.edhi$n) net1lo <- cor1.edlo$cor %>% EBICglasso(., n = cor1.edlo$n) net2lo <- cor2.edlo$cor %>% EBICglasso(., n = cor2.edlo$n) net3lo <- cor3.edlo$cor %>% EBICglasso(., n = cor3.edlo$n) net4lo <- cor4.edlo$cor %>% EBICglasso(., n = cor4.edlo$n) net5lo <- cor5.edlo$cor %>% EBICglasso(., n = cor5.edlo$n) net6lo <- cor6.edlo$cor %>% EBICglasso(., n = cor6.edlo$n) net7lo <- cor7.edlo$cor %>% EBICglasso(., n = cor7.edlo$n) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " 5. Confirmatory Factor analysis To replicate our results without item 1, delete yts_1 from the analyses 5.1 yes/no RESIST  [CODESPLIT] dataCFA1 <- na.omit(data.frame(data$id,data$yts_1, data$yts_2, data$yts_3, data$yts_4,  data$yts_5, data$yts_6,  data$yts_7, data$yts_8,  data$yts_9, data$yts_10, data$yts_11, data$yts_12, data$yts_13)) colnames (dataCFA1) <-c(\"id\",\"yts_1\",\"yts_2\",\"yts_3\",\"yts_4\",\"yts_5\",\"yts_6\", \"yts_7\",\"yts_8\",\"yts_9\",\"yts_10\",\"yts_11\",\"yts_12\", \"yts_13\") model1 <- 'EA =~ yts_1 + yts_2  + yts_3  +  yts_4  + yts_5 + yts_6  + yts_7  +  yts_8  + yts_9 + yts_10 + yts_11 +  yts_12 + yts_13' fit1 <- cfa(model1, data=dataCFA1, meanstructure=T, std.lv=T, estimator = \"WLSMV\", ordered = c(\"yts_1\", \"yts_2\",  \"yts_3\",  \"yts_4\", \"yts_5\", \"yts_6\",  \"yts_7\",  \"yts_8\", \"yts_9\", \"yts_10\", \"yts_11\", \"yts_12\", \"yts_13\")) summary(fit1, fit.measures=TRUE, standardized = T) fitMeasures(fit1,c(\"chisq\",\"df\",\"pvalue\",\"cfi\",\"tli\",\"rmsea\", \"wrmr\",'rmsea.ci.lower','rmsea.ci.upper')) modindices(fit1, sort. = T) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get factorscores and include them in origingal data [CODESPLIT] dataCFA1$factorscore1 <- predict(fit1) data <- left_join(data, dataCFA1,  by = \"id\") data$factorscore1 dataCFA3$factorscore3 <- predict(fit3) data1 <- left_join(data1, dataCFA3,  by = \"record_id\") data1$factorscore3 dataCFA4$factorscore4 <- predict(fit4) data1 <- left_join(data1, dataCFA4,  by = \"record_id\") data1$factorscore4 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "join behDat and demDat [CODESPLIT] allData = inner_join(behDat, demDat, by = \"userCode\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Standardize Variables  zscoring [CODESPLIT] df_overall = df df_overall$sWM = scale(df_overall$sWM) df_overall$rtWM = scale(df_overall$rtWM) df_overall$rtDivAtt = scale(df_overall$rtDivAtt) df_overall$rtAlert = scale(df_overall$rtAlert) df_overall$sRWT = scale(df_overall$sRWT)", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Remove time points above 29, because not everyone attended that often   [CODESPLIT] df_overall = df_overall %>% filter(Time < 29) %>% ungroup()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ensure that userCode is a factor [CODESPLIT] df_overall$userCode = as.factor(df_overall$userCode)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We can also try to cluster the languages based on the feature values they have. We currently use hierarchical clustering. This will normally also reflect the PCA visualization. The output can also be compared with the family grouping of the languages.  generate the clusters [CODESPLIT] hclust_avg <- hclust(daisy(con_data)) hclust_avg <- hclust(daisy(con_data))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "add squares around the clusters [CODESPLIT] rect.hclust(hclust_avg , k = 4, border = 1:4) rect.hclust(hclust_avg , k = 4, border = 1:4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Extracting distances  We can extract the pairwise distance between the points of the data set. First, we can visualize the distances in a twodimensional space. Normally, we expect that this visualization matches the output of the PCA.  change the content to a distance matrix  [CODESPLIT] distances <- con_data %>% dist(method = \"euclidean\") distances <- con_data %>% dist(method = \"euclidean\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "recodes ethnicity (Decline/Other category is set to missing) [CODESPLIT] sambdif<-samb %>% select(1:19, SAmb_tot, gender, ethnic) %>% mutate(ethnic_fac = rec(ethnic, rec = \"1=0;; 6=NA;; else=1\")) %>% convert(fct(gender, ethnic_fac)) table(sambdif$gender) # 0=women, 1=men table(sambdif$ethnic_fac) #0=EuroAm, 1=PoC ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "reorder the list of results to match the alignment of the dendrogram [CODESPLIT] list.res.reo<-list.res[order.dendrogram(dend1)] layd<-c(10,1) tiff(\"Figure_hclust2.tif\",width=sum(layd),height=11.5,units=\"cm\",res=600,compression=\"lzw\") layout(matrix(1:2,nrow=1),widths=layd) par(mar=c(3.5,1,1.5,14),mgp=c(2,0.8,0)) plot(dend1, cex = 0.8, horiz=T,xlab=\"Height\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to read in cleaned data and format  it as a tibble that can be used for plotting [CODESPLIT] read_data_for_plotting <- function(data_path) { dat <- read_csv(data_path, col_types = cols()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a matrix of relevant item loadings to calculate nonrefined factor (scale) scores [CODESPLIT] scaleMatrix <- data.frame(data = matrix(EFA$loadings, ncol = 6)) scaleMatrix <- lapply(scaleMatrix, function(x) ifelse(x >= 0.4, 1, ifelse(x <= -0.4, -1, NA))) scaleMatrix <- data.frame(data = scaleMatrix, row.names = colnames(mainData[, c(2:56)])) colnames(scaleMatrix) <- c(\"Sensory\", \"CognitiveDemand\", \"ThreatToSelf\", \"CrossSettings\", \"Safety\", \"States\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get number of patients (from 'table_overview_of_studies.R') [CODESPLIT] df_long <- left_join(df_long, tab_study_overview %>% select(study_name, N_Patients) %>% rename(Study = study_name) %>% mutate(N_Patients = as.numeric(str_extract(N_Patients, '(\\\\d*)'))), by = 'Study') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "add error bars with standard error [CODESPLIT] plot <- plot + geom_errorbar(aes(ymin=lower,ymax=upper),width=.2) plot <- plot + ggtitle(\"2a) Average accuracy across four Cue Type conditions\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Capture Age for Each particpant [CODESPLIT] ParticipantAge <- summarize(Summary, count = n(), Age = mean(Age, na.rm = T)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "EVENT VALENCE: Compare Rated Strength of Positive & Negative Events    CHECKING FOR OUTLIERS    Figure: Boxplot    Order Conditions [CODESPLIT] EventValenceRatings$Valence <- factor(EventValenceRatings$Valence, levels = c(\"Positive\", \"Negative\")) BoxplotFig <- ggplot(EventValenceRatings, aes(x=Valence, y=Strength)) + theme_bw() BoxplotFig <- BoxplotFig +  geom_boxplot(aes(color = Valence, fill = Valence), outlier.size = 2.5, alpha = 0.5) BoxplotFig <- BoxplotFig + labs(x=\"Event Valence\", y=\"Event Valence Ratings (-50 to 50)\") + theme(title= element_text(size=14, face='bold'), axis.title.x= element_text(size=14), axis.text.x= element_text(face=\"bold\", size=14), axis.title.y = element_text(size=14), axis.text.y= element_text(face=\"bold\", size=14), strip.text.x= element_text(size = 14, face='bold'), strip.text.y= element_text(size = 14, face='bold')) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Dotplot Valence Figure (Winsorized & ReverseCoded) [CODESPLIT] EventValenceRatings_MEANFig <- ggplot(data=EventValenceRatings_ByParticipant_MEAN, aes(x=Valence, y=Strength_ReverseCoded)) EventValenceRatings_MEANFig <- EventValenceRatings_MEANFig + stat_summary(fun=mean, geom=\"bar\", position =\"dodge\", alpha=0.5, aes(fill=Valence)) EventValenceRatings_MEANFig <- EventValenceRatings_MEANFig + theme_bw() EventValenceRatings_MEANFig <- EventValenceRatings_MEANFig + geom_violin(alpha = 0.5, color='grey50') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mixed Effects Model with EventPosition as a Covariate (Linear & Quadratic)  Random Effects Structure simplified to converge:    ChainPosition removed from ChainID    SocialContext removed from Event [CODESPLIT] Analysis1GLMER <- glmer (Present~ SocialContext.C*Valence.C*ChainPosition.C + poly(EventPosition.C, 2) + (1 + Valence.C | ChainID) + (1 | Event), data=Analysis1, family=binomial) summary(Analysis1GLMER) Analysis2GLMER <- glmer (Present~ SocialContext.C*Valence.C*ChainPosition.C + poly(EventPosition.C, 2) + (1 + Valence.C | ChainID) + (1 | Event), data=Analysis2, family=binomial) summary(Analysis2GLMER) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mixed Effects Model with EventPosition as a Covariate (Linear & Quadratic)  Random Effects Structure simplified to converge:    Valence & ChainPosition removed from ChainID    SocialContext removed from Event [CODESPLIT] Analysis3GLMER <- glmer (Present~ SocialContext.C*Valence.C*ChainPosition.C + poly(EventPosition.C, 2) + (1 | ChainID) + (1 | Event), data=Analysis3, family=binomial) summary(Analysis3GLMER) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "   VISUALISE BIAS (POSITIVE, NEGATIVE) IN EACH SocialContext      Substract Mean Positive From Mean Negative Survival Score for each SocialContext  Prediction is that Perference for Negativity Decreases first with Communicative Intent and then further with Social Interaction  Calculate Means for Positive & Negative Information  Calculate Mean Positive  Mean Negative for Each Chain  Plot  Filter  [CODESPLIT] Bias <- filter(NegativityData, Valence %in% c(\"Positive\", \"Negative\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a Vector of colours based on each Participant's Mean Bias Score (Positive  Negative)  Use this to Color the dot points by Value [CODESPLIT] wideBiasMEAN <- mutate(wideBiasMEAN, ColourFig = ifelse(Bias >1, \"#006A40FF\", ifelse(Bias <0, \"#F08892FF\", \"#95828DFF\"))) NegativityBiasFig <- ggplot(data=wideBiasMEAN, aes(x=SocialContext, y=Bias)) NegativityBiasFig <- NegativityBiasFig + stat_summary(fun.y=mean, geom=\"bar\", position =\"dodge\", alpha=0.5, fill=\"red\") NegativityBiasFig <- NegativityBiasFig + theme_bw() NegativityBiasFig <- NegativityBiasFig + geom_violin(alpha = 0.5, color='grey50') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Geom_dotplot with points coloured by Value [CODESPLIT] NegativityBiasFig <- NegativityBiasFig + geom_dotplot(aes(x=SocialContext, y=Bias, fill=ColourFig), binaxis='y', stackdir='center', dotsize=0.9, alpha = 1, stroke=0.75, colour='grey15') + scale_fill_identity() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add Violin Geom to give indication or data normality [CODESPLIT] ResolutionCollapsedFig <- ResolutionCollapsedFig + geom_violin(alpha = 0.0, colour='grey50') ResolutionALLPosNegFig <- ResolutionALLPosNegFig + geom_violin(alpha = 0.5, color='grey50') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Run IRT model: 3dim model for math, read  and scie, no weights  Possibility to set seeds and iteration number in the function parameter  [CODESPLIT] mod <- run.irt(pa12_resp_s, items = item.dif) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to create tick marks on a logscale [CODESPLIT] log10Tck <- function(side, type){ lim <- switch(side, x = par('usr')[1:2], y = par('usr')[3:4], stop('side argument must be \"x\" or \"y\"')) at <- floor(lim[1]) : ceiling(lim[2]) at <- 0:8 return(switch(type, minor = outer(1:9, 10^(min(at):max(at)))[,1:8], major = 10^at, stop('type argument must be \"major\" or \"minor\"') )) } if(saveFigures) cairo_ps(file = '../R_Output/Images/OriginalBFScatter.eps', onefile = TRUE, fallback_resolution = 600, width = 7.3, height = 4.16) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot: original Bayes factor and replication effect size [CODESPLIT] plot(original.bf, replication.effectsizes, xlim    = c(1, 10^8), ylim    = c(-0.2, 1), axes    = FALSE, log     = 'x', cex     = 2, cex.lab = 1.6, pch     = 21, bg      = c('grey36', 'grey')[replication.outcomes], xlab    = 'Bayes Factor Original Study', ylab    = 'Replication Effect Size (r)') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " take the xaxis indices and add a jitter, proportional to the N in each level  [CODESPLIT] myjitter <- jitter(rep(i, length(thisvalues)), amount=levelProportions[i]/6) points(myjitter, thisvalues, pch=1, col=rgb(0,0,0,.9), cex = 1.5, lwd = 1.5) } graphics.off() rm(myjitter, thislevel, thisvalues, mylevels, levelProportions, nsubjects, subjects) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Multiple Regression Analysis [CODESPLIT] multreg.second(Speaking ~ Vocabulary+Grammar+Writing+Reading, corr=correl, n=100) multreg.second(Y~ X1+X2+X3, corr=correl, n=100) multreg.second(Score~ Wordcount+CLI+Commas+Stopwords+Linking+WordsSentence, corr=correl, n=200) multreg(Score~ Wordcount+CLI+Commas+Stopwords+Linking+WordsSentence, data=dat1) lm.out <- lm(Score ~., dat1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Dominance Analysis (using dominanceanalysis package) [CODESPLIT] lm.cov <- lmWithCov(Score ~ Wordcount+CLI+Commas+Stopwords+Linking+WordsSentence, correl) da <- dominanceAnalysis(lm.cov) print(da) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "FUnctions for Cleaning   Function to generate a table with all observations within a variable and its corresponding count [CODESPLIT] check_observation <- function (df, column) { require (tidyverse) check <- df %>% group_by (as.vector(unlist(df[, column]))) %>% count () colnames(check)[1] <- column return (check) } recur.collapse <- function(df, n = ncol(df)) { #collapse function combine <- function(x,y){ if(!is.na(x)){ return(x) } else if (!is.na(y)) { return(y) } else { return(NA) } } if(n==1){ return(df[,1]) } else { return(mapply(combine, df[,n], recur.collapse(df, n-1))) } } multi_spread <- function(df, id, key, var){ require(gtools) require(tidyverse) df.list <- lapply(var, function(x){ df.temp <- spread(df[, c(id, key, x)], key, x) # spread 1 variable cname <- mixedsort(unique(as.vector(unlist(df[,key])))) # sort colnames in alphanumeric order df.temp <- df.temp[, c(id, cname)] # rename column names colnames(df.temp)[2:ncol(df.temp)]  <- paste(x, colnames(df.temp[,cname]), sep= \"_\") #add key to column names return(df.temp) }) df.output <- df.list[[1]] # initialise for(i in 1:(length(var)-1)){ # combine matrices together df.output <- full_join(df.output, df.list[[i+1]]) } return(as.data.frame(df.output)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Extract posteriors of effects from Bayesian ANOVA [CODESPLIT] bestmod <- modBF[1] chains <- posterior(bestmod, iterations=50000, columnFilter=\"^Subject$\") colnames(chains) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "STATISTICAL TEST: Nonparametric Spearman's correlation nonnormalized conditional entropy [CODESPLIT] cor.test(results$DATE, results$CEDenominationsMotifs, method = \"spearman\") rdates <- rev(dates) #dates = centuries BCE plot(results$DATE, results$CEDenominationsMotifs, xlim = c(6,4), xaxt='n', xlab = \"Century BCE\", ylab = \"H(D|d)\", main = \"P2: Conditional entropy of denominantions given designs\") axis(1, at = rdates, labels = rdates) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "____________________________________________________________ Plots with H(D|d)  nonnormalized conditional entropy Getting statistics: mean, meadian, standard deviation among authorities per period for nonnormalized conditional entropy of denomination given designs  [CODESPLIT] N <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = length) MEAN <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = mean) MEDIAN <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = median) SD <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = sd) resultspoleis_summary <- cbind.data.frame(N, MEAN$CEDenominationsMotifs, MEDIAN$CEDenominationsMotifs, SD$CEDenominationsMotifs) colnames(resultspoleis_summary) <- c(\"DATE\",\"N\",\"MEAN\",\"MEDIAN\",\"SD\") resultspoleis_summary$SE <- resultspoleis_summary$SD / sqrt(resultspoleis_summary$N) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Getting statistics: mean, meadian, standard deviation among authorities per period for normalized conditional entropy of denomination given designs  [CODESPLIT] N <- aggregate(NormCEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = length) MEAN <- aggregate(NormCEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = mean) MEDIAN <- aggregate(NormCEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = median) SD <- aggregate(NormCEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = sd) Nresultspoleis_summary <- cbind.data.frame(N, MEAN$NormCEDenominationsMotifs, MEDIAN$NormCEDenominationsMotifs, SD$NormCEDenominationsMotifs) colnames(Nresultspoleis_summary) <- c(\"DATE\",\"N\",\"MEAN\",\"MEDIAN\",\"SD\") Nresultspoleis_summary$SE <- Nresultspoleis_summary$SD / sqrt(Nresultspoleis_summary$N) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " aggregate data, zstandardize all predictor variables for the model  [CODESPLIT] recogfreq <- aggregate(cbind(correct, intrusion, new) ~ id+setsize+rsizeList+rsizeNPL, data=recogdat, FUN=sum) recogfreq$zsetsize <- (recogfreq$setsize - mean(recogfreq$setsize))/sd(recogfreq$setsize) recogfreq$zrsizeList <- (recogfreq$rsizeList - mean(recogfreq$rsizeList))/sd(recogfreq$rsizeList) recogfreq$zrsizeNPL <- (recogfreq$rsizeNPL - mean(recogfreq$rsizeNPL))/sd(recogfreq$rsizeNPL) for (sc in 1:length(scales))  { prior <- paste0(\"cauchy(0, \", as.character(scales[sc]), \")\") fixefPrior <- c(set_prior(prior, class=\"b\")) ranefPrior <- set_prior(\"gamma(1,0.04)\", class=\"sd\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Cronbach's alpha (ingroup) [CODESPLIT] iden.ing.woOutliers.a <- c(\"ingroup.value\", \"ingroup.like\", \"ingroup.connected\") iden.ing.woOutliers <- data.all[iden.ing.woOutliers.a] cronbach(iden.ing.woOutliers) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " min and max age, mean and sd age, percentage of men and women  [CODESPLIT] minAge = min(df_preQ$age) maxAge = max(df_preQ$age) meanAge = mean(df_preQ$age) sdAge = sd(df_preQ$age) females = length(which(df_preQ$gender == \"female\")) males = length(which(df_preQ$gender == \"male\")) other = length(which(df_preQ$gender == \"other\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Check normality with QQ plot and ShapiroWilk test  Build the linear model [CODESPLIT] model  <- lm(FW ~ condition, data = df_postQ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Question 4: pearson correlation between the different subscales [CODESPLIT] df_cor <- data.frame(df_postQ$FW, df_postQ$DU, df_postQ$DET) colnames(df_cor) <- c('FW', 'DU', 'DET') res_cor <- rcorr(as.matrix(df_cor)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Visualize correlations  Insignificant correlations are leaved blank [CODESPLIT] corrplot(res_cor$r, method = 'number', type=\"upper\", order=\"hclust\", p.mat = res_cor$P, sig.level = 0.05, insig = \"blank\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Stepwise regression model [CODESPLIT] step.model <- stepAIC(full.model, direction = \"both\", trace = FALSE) summary(step.model) step.model <- stepAIC(full.model, direction = \"both\", trace = FALSE) summary(step.model) step.model <- stepAIC(full.model, direction = \"both\", trace = FALSE) summary(step.model) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Subset the data set to use base2 instead of base1 for the individuals  that have the 2nd baseline  Extract values of frogs that have two baselines [CODESPLIT] base2_frogs <-filter(hormones, condition == \"base_02\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "LMER testosterone by sex and condition with interaction [CODESPLIT] m1_osT <- lmer(log(t_conc_corr) ~ sex*condition + (1|id), data = hormones) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Sex difference in Tlevel  Model plot wit sjPlot [CODESPLIT] plot_model(m1_osT_1, title = \"\", axis.title = \"Fixed effect estimates\", axis.labels = c(\"Time point\\n (back home)\", \"Sex \\n(male)\"), dot.size = 5, line.size = 2, transform = NULL, sort.est = TRUE, colors = \"gs\", vline.color = \"darkgrey\")+ ylim(-0.5, 1) + theme_sjplot2(24) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "now the random slopes part of the model including random intercepts and slopes and their correlation: [CODESPLIT] pot.terms.with.corr=paste(c(xnames, fe.me[modes[fe.me]!=\"factor\"]), collapse=\"+\")#get all fixed effects terms together... pot.terms.with.corr=paste(paste(\"(1+\", pot.terms.with.corr, \"|\", re, \")\", sep=\"\"), collapse=\"+\")#... and paste random effects, brackets and all that pot.terms.with.corr=paste(c(xnames, fe.me[modes[fe.me]!=\"factor\"]), collapse=\"+\")#get all fixed effects terms together... pot.terms.with.corr=paste(paste(\"(1+\", pot.terms.with.corr, \"|\", re, \")\", sep=\"\"), collapse=\"+\")#... and paste random effects, brackets and all that ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Supplemental Material:  To check whether the participants followed gaze at all, we tested the gaze following score against chance level by running a one sample test against zero.  Children followed gaze in the additional gazefollowing task (M  3;; SD  1.66;; t(29)  9.89, p  .00, d  1.8).  Correlation GF & Difference Score  [CODESPLIT] Data$DiffSocCon <- as.numeric(Data$DiffSocCon) cor.test(Data$DS, Data$DiffSocCon, method = \"pearson\") #only younger ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "use network data only for students in classrooms with at least 50% response rate [CODESPLIT] dat$nwinclude <- 0 dat[which(dat$prop_part > .50 & !is.na(dat$prop_part)),\"nwinclude\"] <- 1 length(unique(dat[which(dat$nwinclude == 1), \"IDTESTGROUP_FDZ\"]))    # 1708 classes with at least 50% response rate length(unique(dat[which(dat$nwinclude == 0), \"IDTESTGROUP_FDZ\"]))    # 290 classes excluded for network related analyses dat[which(dat$nwinclude == 0), c(nw_covariates,  grep(\"deg\", names(dat), value = T))] <- NA #recode sociometric variables to NA for classes with low level of completeness length(unique(dat[which(dat$nwinclude == 0), \"IDSTUD_FDZ\"]))         # 5450 students without sociometric data length(unique(dat[which(dat$nwinclude == 1), \"IDSTUD_FDZ\"]))         # 36,920 students with sociometric data ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Reorder factor levels: (this will be useful once we create the plot legend) [CODESPLIT] dataset$trial_condition <- factor(dataset$trial_condition, levels = c(\"diff_3sg\", \"diff_3pl\", \"same_3sg\", \"same_3pl\")) dataset$trial_condition_new <- factor(dataset$trial_condition_new, levels = c(\"different number\", \"same number\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Prepare dataframe for eyetracking analysis & graphs (package eyetrackingR required) [CODESPLIT] eyetrackingr.data <- make_eyetrackingr_data(dataset, participant_column = \"participant_number\", trial_column = \"item_nr\", time_column = \"time_ms_absolute\", trackloss_column = \"trackloss\", aoi_columns = c(\"average_target_sample_count_proportion\", \"average_distractor_sample_count_proportion\"), treat_non_aoi_looks_as_missing = TRUE ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Remove trackloss per trial > 50% (removed 4 trials) [CODESPLIT] eyetrackingr.data <- clean_by_trackloss(data = eyetrackingr.data, trial_prop_thresh = .5) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check histogram for binning [CODESPLIT] hist(quop_use$t1_eff) quop_use$bins<-cut(quop_use$t1_eff,quantile(quop_use$t1_eff,p=seq(0,1,length=29),na.rm=T)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "visualize with forest plot  compare corona vs. precorona [CODESPLIT] forest(lvd$est[1:16],sei=lvd$se[1:16], ilab = cbind(c(\"Pre-pandemic Cohorts vs. First Pandemic Cohort\",rep(\"\",7),\"Pre-pandemic Cohorts vs. Second Pandemic Cohort\",rep(\"\",7)), paste0(\"T\",rep(1:8,2)), c(\"Schools Open\",\"???\",\"???\",\"???\",\"Schools Closed\",\"???\",\"???\",\"Alternating Lessons\", \"Schools Open\",\"???\",\"Alternating Lessons\",\"???\",\"???\",\"???\",\"???\",\"Schools Open\")), ilab.xpos = c(-1.35,-.95,-.65), ylim = c(0,22), xlim = c(-1.7,1), slab = NA, rows = c(18:11, 8:1), at = c(-.7,-.35,0,.35,.7), xlab = \"Latent Variance Differences\", col=c(\"gray\",\"gray\", \"gray\",\"gray\", \"black\", \"black\", \"black\", \"black\", \"gray\", \"gray\", \"black\", \"black\", \"black\", \"black\", \"black\", \"gray\")) forest(lmd$est[1:16],sei=lmd$se[1:16], ilab = cbind(c(\"Pre-pandemic Cohorts vs. First Pandemic Cohort\",rep(\"\",7),\"Pre-pandemic Cohorts vs. Second Pandemic Cohort\",rep(\"\",7)), paste0(\"T\",rep(1:8,2)), c(\"Schools Open\",\"???\",\"???\",\"???\",\"Schools Closed\",\"???\",\"???\",\"Alternating Lessons\", \"Schools Open\",\"???\",\"Alternating Lessons\",\"???\",\"???\",\"???\",\"???\",\"Schools Open\")), ilab.xpos = c(-1.35,-.95,-.65), ylim = c(0,22), xlim = c(-1.7,1), slab = NA, rows = c(18:11, 8:1), at = c(-.7,-.35,0,.35,.7), xlab = \"Standardized Latent Mean Differences\", col=c(\"gray\",\"gray\", \"gray\",\"gray\", \"black\", \"black\", \"black\", \"black\", \"gray\", \"gray\", \"black\", \"black\", \"black\", \"black\", \"black\", \"gray\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "to see how well the distances are measured [CODESPLIT] stressplot(species_matrix_short.nmds) #R2 = 0.961, linear R2 = 0.891 data.scores <- as.data.frame(scores(species_matrix_short.nmds$points)) data.scores$site <- substr(rownames(data.scores),1,6) data.scores$plots <- substr(rownames(data.scores),8,15) species.scores <- as.data.frame(scores(species_matrix_short.nmds, \"species\")) species.scores$species <- rownames(species.scores) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "now statistical test to see if communities are statistically different from one another [CODESPLIT] species_matrix2 <- species_matrix %>% mutate_if(is.numeric, ~1 * (. > 0)) sp_matrix_with_group <- species_matrix2 rownames(sp_matrix_with_group) <- rownames(species_matrix) sp_matrix_with_group$grouping <- substr(rownames(sp_matrix_with_group), 7,13) ano <- anosim(species_matrix2, distance = 'euclidean', grouping = sp_matrix_with_group$grouping) summary(ano) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " SEM Analysis Section   The following section sets up a series of path models. Note that this is a highly revised version of the SEM section, which  originally tried to incorporate a complex latent variable model to simultaneously test all hypotheses. I decided to break it down into  smaller, simpler models, and to focused on observedvariableonly models using the composite meausres instead. I'm not sure if that's  the \"best\" approach or not, but I think it at least makes for a simpler approach. The code below is thus a mixture of what I originally  wrote and the revised models.  Let's check differences between the two groups:  [CODESPLIT] t.test(Measures$Age~Measures$Normative) t.test(Measures$Adj_Income~Measures$Normative) t.test(Measures$Parent_Edu~Measures$Normative) t.test(Measures$Political~Measures$Normative) t.test(Measures$Crit_Reflection_Mean~Measures$Normative) t.test(Measures$Efficacy_Mean~Measures$Normative) t.test(Measures$Crit_Action_Mean~Measures$Normative) t.test(Measures$BLM_Activism~Measures$Normative) t.test(Measures$Militarism_Mean~Measures$Normative) t.test(Measures$EDNhComp~Measures$Normative) t.test(Measures$ACEs_Sum~Measures$Normative) t.test(Measures$MEQ_Total~Measures$Normative) t.test(Measures$Discrim_Major_Sum~Measures$Normative) t.test(Measures$Discrim_Everyday_Mean~Measures$Normative) t.test(Measures$BNSS_Mean~Measures$Normative) t.test(Measures$BNSSh_Mean~Measures$Normative) t.test(Measures$Zero_Sum_Mean~Measures$Normative) sjmisc::frq(Measures$Race, out = \"v\") Measures %>% group_by(Race) %>% summarise(BLM_Mean = mean(BLM_Activism)) data <- Measures %>% filter(Race %in% c(\"Black or African-American\",\"White or Caucasian\") & !is.na(BLM_Activism)) %>% mutate(Black = ifelse(Race == \"Black or African-American\", 1,0)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " To assess BLM Activism by race, keep in mind that the BLM measure is count data (with a lot of zeros)  A ttest therefore is not appropriate. We will use \"zeroinflated Poisson regression\"  [CODESPLIT] ggplot(Measures, aes(BLM_Activism)) + geom_histogram() model.zi = zeroinfl(BLM_Activism ~ Black, data = data, dist = \"poisson\") summary(model.zi) Descriptives3 <- Measures %>% select(Normative,Age,Adj_Income,Parent_Edu,Political,ACEs_Sum,EDNhComp,Discrim_Major_Sum,Discrim_Everyday_Mean,MEXQ_Exp, Crit_Reflection_Mean,Efficacy_Mean,Crit_Action_Mean,BLM_Activism, Militarism_Mean) %>% psych::describeBy(group=\"Normative\") %>% as.data.frame() %>% rownames_to_column(var=\"Measure\") %>% select(-vars) DescrNonNorm<- as.data.frame(Descriptives3[1]) %>% rownames_to_column(var=\"Measure\") DescrNorm<- as.data.frame(Descriptives3[2]) %>% rownames_to_column(var=\"Measure\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Write cleaned covariate files to xlsx [CODESPLIT] cleaned_group_covariates_data <- list( \"cleaned_group_participant\" = cleaned_group_participant_dat, \"cleaned_group_instructional\" = cleaned_group_instructional_dat, \"cleaned_group_DV\" = cleaned_group_DV_dat, \"cleaned_group_race_ethnicity\" = race_eth ) write_xlsx(cleaned_group_covariates_data, path = \"Cleaned data/Group covariates data tables.xlsx\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Weighted regression model for Ordinary People subscale score (mean) [CODESPLIT] m1.ppl.svyglm.fit <- svyglm(scipopppl ~ age + gender + education.comp + education.uni + sciprox.score + urbanity.log + languageregion.ger + languageregion.ita, design = bar.design.scipopppl, family = gaussian, na.action = na.omit) m2.ppl.svyglm.fit <- svyglm(scipopppl ~ age + gender + education.comp + education.uni + sciprox.score + urbanity.log + languageregion.ger + languageregion.ita + polorientation + religiosity, design = bar.design.scipopppl, family = gaussian, na.action = na.omit) m3.ppl.svyglm.fit <- svyglm(scipopppl ~ age + gender + education.comp + education.uni + sciprox.score + urbanity.log + languageregion.ger + languageregion.ita + polorientation + religiosity + interestscience + sciliteracy + trustscience + trustscientists, design = bar.design.scipopppl, family = gaussian, na.action = na.omit) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Model assumption checks   Do some assumption checks for multiple linear regression (e.g., see Field, 2012, p. 292). These are:  (1) Multicollinearity  (2) Nonnormality/heteroscedasticity of residuals  Assumption checks (1): Multicollinearity   Specify model and inspect GVIFs  Note: The VIF (usual collinearity diagnostic) may not be applicable to models with dummy regressors  constructed from a polytomous categorical variable or polynomial regressors (Fox, 2016: 357). Fox and  Monette (1992) introduced generalized variance inflation factor (GVIF) for these cases. As  education.uni/education.comp and languageregion.ger/languageregion.ita are such dummy regressors based  on polytomous categorial variables (i.e. education and languageregion), we should compute GVIFs.  car::vif() does that automatically.  GVIFs of SciPop Score models  [CODESPLIT] car::vif(m1.scipopgoertz.svyglm.fit) %>% as.data.frame %>% rename(GVIF = deparse(substitute(.))) car::vif(m2.scipopgoertz.svyglm.fit) %>% as.data.frame %>% rename(GVIF = deparse(substitute(.))) car::vif(m3.scipopgoertz.svyglm.fit) %>% as.data.frame %>% rename(GVIF = deparse(substitute(.)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Nonnormality/heteroscedasticity of residuals of ppl Score models [CODESPLIT] distributionchecks(m1.ppl.svyglm.fit) distributionchecks(m2.ppl.svyglm.fit) distributionchecks(m3.ppl.svyglm.fit)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ADDITIONAL ANALYSIS B: TESTING A USHAPED RELATIONSHIP BETWEEN SCIPOP AND POLITICAL ORIENTATION   1st test: Quadratic regression on weighted data [CODESPLIT] bar2019pop$polorientation2 <- bar2019pop$polorientation^2 bar.design <- svydesign(id = ~0, data = bar2019pop, weights = ~weight) quad.scipopgoertz.svyglm.fit <- svyglm(scipopgoertz ~ age + gender + education.comp + education.uni + sciprox.score + urbanity.log + languageregion.ger + languageregion.ita + polorientation + polorientation2 + religiosity + interestscience + sciliteracy + trustscience + trustscientists, design = bar.design, family = gaussian, na.action = na.omit) summ(model = quad.scipopgoertz.svyglm.fit, binary.inputs = \"full\", n.sd = 2, transform.response = F, confint = F, vifs = F, scale = F, model.info = T,  model.fit = T, digits = 4) quad.scipopgoertz.svyglm.fit$aic m0.scipopgoertz.svyglm.fit <- svyglm(scipopgoertz ~ 1, bar.design, family = gaussian, na.action = na.omit) anova(m0.scipopgoertz.svyglm.fit, quad.scipopgoertz.svyglm.fit, test = \"F\", method = \"Wald\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "chisquare to compare vegan in experienced vs. not experienced for each advocacy type graphic video [CODESPLIT] data %>% xtabs(~ graphic_exp_buc + vegn_bin, .) %>% print() %>% proportions(\"graphic_exp_buc\") graphic_tmp <- chisq.test(xtabs(~ graphic_exp_buc + vegn_bin, data)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Boosted Regression Trees                          CUS BRT plot [CODESPLIT] CUS.BRT.plot <- ggplot(data = CUS.BRT, aes(x = predictors, y = influence, fill = predictor.type)) + geom_bar(stat = \"identity\") + geom_hline(yintercept = 5, linetype = 3, size = 1, colour = \"gray60\") + scale_fill_manual(values = viridis.3) + coord_flip() + labs(x = \"Predictor\", y = \"Relative Influence\") + scale_y_continuous(breaks = c(0, 10, 20, 30, 40), limits = c(0, 40)) + scale_x_discrete(limits = BRT.plot.label.limits, labels = BRT.plot.labels) + theme_pubr() + theme(legend.position = c(0.85, 0.825), legend.title = element_blank()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Sets limits of the plot based on user choice [CODESPLIT] limits <- matrix(0,2,2) limits <- if (lim == 1){ limits <-u.plot.limit(min.age,max.age,lim,min.ylim,max.ylim) } else if (lim == 2){ limits <-u.plot.limit(min.age,max.age,lim,min.ylim,max.ylim) } else { limits <-u.plot.limit(min.xlim,max.xlim,lim,min.ylim,max.ylim) } limconc <-uconcplot(limits[1,1],limits[1,2],int) #sets the concordia line based on plot limts plot_limconc <- limconc[2:(nrow(limconc)-1),] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "BELOW CALCULATES THE ELLIPSES               calculate x/y coords that define ellipse based in input data including rho [CODESPLIT] ell.coords=list() #sets up list to write results to for(n in 1:n){ covmat <-cor2cov2((Pb7Ue[n]/2),(Pb6Ue[n]/2),rho[n]) nn <- 75 cutoff <- stats::qchisq(1-0.05,2) e <- eigen(covmat) a <- sqrt(cutoff*abs(e$values[1])) # major axis b <- sqrt(cutoff*abs(e$values[2])) # minor axis v <- e$vectors[,1] beta <- atan(v[2]/v[1]) theta <- seq(0, 2 * pi, length=nn) out <- matrix(0,nrow=nn,ncol=3) out[,1] <- Pb7U[n] + a * cos(theta) * cos(beta) - b * sin(theta) * sin(beta) out[,2] <- Pb6U[n] + a * cos(theta) * sin(beta) + b * sin(theta) * cos(beta) out[,3] <- n ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Functions for plotting models Extracts significance from models for chart legend [CODESPLIT] extractSign <- function (model, modeltype, MtnRanges, InclMIA = TRUE) { coef <- summary(model)$coefficients$cond SignDF <- data.frame(mtnrng = MtnRanges, main = NA, int_ao = NA, int_po = NA) if (InclMIA) { if (modeltype == \"int\") { if (coef[row.names(coef) == \"AO:MtnRange1\", 4] < 0.05) {SignDF$int_ao[SignDF$mtnrng == \"Coast - North\"] <- \"I\"} if (coef[row.names(coef) == \"AO:MtnRange2\", 4] < 0.05) {SignDF$int_ao[SignDF$mtnrng == \"Coast - South\"] <- \"I\"} if (coef[row.names(coef) == \"AO:MtnRange3\", 4] < 0.05) {SignDF$int_ao[SignDF$mtnrng == \"Columbia - North\"] <- \"I\"} if (coef[row.names(coef) == \"AO:MtnRange4\", 4] < 0.05) {SignDF$int_ao[SignDF$mtnrng == \"Columbia - South\"] <- \"I\"} if (coef[row.names(coef) == \"AO:MtnRange5\", 4] < 0.05) {SignDF$int_ao[SignDF$mtnrng == \"Rockies - North\"] <- \"I\"} if (coef[row.names(coef) == \"MtnRange1:PO\", 4] < 0.05) {SignDF$int_po[SignDF$mtnrng == \"Coast - North\"] <- \"I\"} if (coef[row.names(coef) == \"MtnRange2:PO\", 4] < 0.05) {SignDF$int_po[SignDF$mtnrng == \"Coast - South\"] <- \"I\"} if (coef[row.names(coef) == \"MtnRange3:PO\", 4] < 0.05) {SignDF$int_po[SignDF$mtnrng == \"Columbia - North\"] <- \"I\"} if (coef[row.names(coef) == \"MtnRange4:PO\", 4] < 0.05) {SignDF$int_po[SignDF$mtnrng == \"Columbia - South\"] <- \"I\"} if (coef[row.names(coef) == \"MtnRange5:PO\", 4] < 0.05) {SignDF$int_po[SignDF$mtnrng == \"Rockies - North\"] <- \"I\"} } } SignAOMain <- paste(\"AO: \", format(round(coef[row.names(coef) == \"AO\", 1], 3), nsmall=3), \" (\", format(round(coef[row.names(coef) == \"AO\", 4], 3), nsmall=3),\")\") SignPOMain <- paste(\"PO: \", format(round(coef[row.names(coef) == \"PO\", 1], 3), nsmall=3), \" (\", format(round(coef[row.names(coef) == \"PO\", 4], 3), nsmall=3),\")\") return(list(SignAOMain = SignAOMain, SignPOMain = SignPOMain, SignDF = SignDF)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "outlier coding M and SD per subject and item depending on distractor condition and soa (independent variables) [CODESPLIT] subj.M.SD = ddply(rawdat,.(subj,cond,soa), summarize, subj.M=mean(RT, na.rm=T), subj.SD=sd(RT, na.rm=T)) tar.M.SD = ddply(rawdat,.(targ_ID,cond,soa), summarize, tar.M=mean(RT, na.rm=T), tar.SD=sd(RT, na.rm=T)) rawdat = merge(rawdat, subj.M.SD, by=c(\"subj\", \"cond\", \"soa\")) rawdat = merge(rawdat, tar.M.SD, by=c(\"targ_ID\", \"cond\", \"soa\")) rawdat$subj.min = (rawdat$subj.M - 2*(rawdat$subj.SD)) rawdat$subj.max = (rawdat$subj.M + 2*(rawdat$subj.SD)) rawdat$tar.min = (rawdat$tar.M - 2*(rawdat$tar.SD)) rawdat$tar.max = (rawdat$tar.M + 2*(rawdat$tar.SD)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate the value of the closest peer estimate and store in 'closest' [CODESPLIT] closest<-min(abs(si-firstEstimate))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "define Gaussians for updating  prior self (firstEstimate) [CODESPLIT] density_self <- log(dnorm(x, firstEstimate, own_sd )) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1.1. Sample 1  PTSD Symptoms (PSSI)   Run linear mixed effect model [CODESPLIT] a1_model_01_sg_pssi_outcome <- lme(pssi_end ~ pssi_s0 + sg, random = ~ 1 | id, method = \"ML\", na.action = na.omit, data = data_a1_pssi) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Check asssumption of normality of the residuals [CODESPLIT] qqnorm(resid(a1_model_01_sg_pssi_outcome)) qqnorm(resid(a2_model_01_sg_pssi_outcome))", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Calculate contrasts specified in in the contrast matrix \"contrast_outcome\"  [CODESPLIT] a1_model_01_sg_pssi_outcome_contrasts <- glht(a1_model_01_sg_pssi_outcome, contrast_outcome) a2_model_01_sg_pssi_outcome_contrasts <- glht(a2_model_01_sg_pssi_outcome, contrast_outcome) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Show contrasts without adjusting pvalues for multiple comparisons [CODESPLIT] summary(a1_model_01_sg_pssi_outcome_contrasts, test = adjusted(\"none\")) summary(a2_model_01_sg_pssi_outcome_contrasts, test = adjusted(\"none\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Select estimates, standard error and pvalue  [CODESPLIT] table_a1_model_01_sg_pssi_outcome_full <- table_a1_model_01_sg_pssi_outcome_raw %>% dplyr::select(beta = estimate, se = std.error, p = p.value) table_a2_model_01_sg_pssi_outcome_full <- table_a2_model_01_sg_pssi_outcome_raw %>% dplyr::select(beta = estimate, se = std.error, p = p.value) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate pooled standard deviation at baseline [CODESPLIT] sd_a1_pssi_s0 <- sd(data_a1_pssi$pssi_s0, na.rm = TRUE) sd_a2_pssi_s0 <- sd(data_a2_pssi$pssi_s0, na.rm = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate Cohen's d using the absolute value of the baseline adjusted difference from the linear mixed effect model [CODESPLIT] cohens_d_a1_pssi_end <- a1_pssi_end_diff / sd_a1_pssi_s0 cohens_d_a2_pssi_end <- a2_pssi_end_diff / sd_a2_pssi_s0", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Peer Rank Means and Standard Deviations of Conditions   Unspecified Peer Rank [CODESPLIT] round(mean(unspecified$peer.rank), 2) round(sd(unspecified$peer.rank), 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "determines and rounds (to digits as specified in digits) mean and sd of x, and then collapses them in a single entry  [CODESPLIT] xx=as.character(round(c(mean(x, na.rm=T), sd(x, na.rm=T)), digits=digits)) if(any(grepl(x=xx, pattern=\".\", fixed=T))>0){ xx[!grepl(x=xx, pattern=\".\", fixed=T)]=paste(xx[!grepl(x=xx, pattern=\".\", fixed=T)], \"0\", sep=\".\") xx=matrix(unlist(strsplit(as.character(xx), split=\".\", fixed=T)), ncol=2, byrow=T) xx[, 2]=unlist(lapply(xx[, 2], function(x){ paste(c(x, paste(c(rep(\"0\", times=digits-nchar(x))), collapse=\"\")), collapse=\"\") })) xx=apply(xx, 1, paste, collapse=\".\") } paste(xx, collapse=sep) } c.tab<-function(x, digits=NA, n.spaces=1, add.hash=T, incl.fst=F, incl.rownames=T){ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "wrapper for savePlot with file  \"clipboard\" and type  \"wmf\" (default;; others are possible)  [CODESPLIT] savePlot(file=\"clipboard\", type=type) } overdisp.correction<-function(coeffs, disp.param){ coeffs[, \"Std. Error\"]=coeffs[, \"Std. Error\"]*sqrt(disp.param) coeffs[, \"z value\"]=coeffs[, \"Estimate\"]/coeffs[, \"Std. Error\"] coeffs[, \"Pr(>|z|)\"]=2*pnorm(q=-abs(coeffs[, \"z value\"]), mean=0, sd=1) return(coeffs) } merge.ests.ci.stab<-function(coeffs, ci=NULL, tests=NULL, stab=NULL){ ires=coeffs if(!is.null(ci)){ coeffs=cbind(coeffs, ci[rownames(coeffs), ]) } if(!is.null(tests)){ xx=outer(rownames(tests), rownames(coeffs), Vectorize(function(tt, e){ if(nchar(tt)>nchar(e)){ return(0) }else{ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add the starttime (intra_scope_window[1]) to every start position (scope$start) ...  ... of the recording timestamp to get the actual start window in milliseconds [CODESPLIT] starting_times <- df$RecordingTimestamp[scope$start] + as.numeric(intra_scope_window[1]) } if (intra_scope_window[2] != \"end\") { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "set to current trial duration to 0 [CODESPLIT] current_trial_total_duration[[hn]] <- 0 } current_trial_total_looks[[hn]] <- 0 } current_first_look_duration[[hn]] <- 0 current_first_look_ending_reason[[hn]] <- \"\" first_looks_collection[[hn]]$found_first <- FALSE first_looks_collection[[hn]]$forced_stop <- FALSE } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Return sum log likelihood. Include protection against log(0) problems [CODESPLIT] return(sum(log(pmax(like, 1e-10)))) } else { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot posterior fit and forward simulation prediction [CODESPLIT] plot_fits <- function(stan_fit_ex, N_t, N_p, obs_times, pred_times, data_vector, S_0, D_0, M_0, filename) { sigma_hat <- median(stan_fit_ex$sigma) CO2_flux_ratios_hat_median <- rep(NA, N_t) #Pre-allocate vector for median model fit CO2_flux_ratios_pred_median <- rep(NA, N_p) sigma_hat <- median(stan_fit_ex$sigma) for (t in 1:N_t) { CO2_flux_ratios_hat_median[t] <- median(stan_fit_ex$CO2_flux_ratios_hat_vector[,t]) } for (t in 1:N_p) { CO2_flux_ratios_pred_median[t] <- median(stan_fit_ex$CO2_flux_ratios_new_vector[,t]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit linear mixed effects models  null model [CODESPLIT] fitlmer0 <- lmer(rating ~ 1 + (1|subject), data = dat.long1, REML = FALSE) summary(fitlmer0) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "the mean of each parameter across iterations. Keep dimensions for parameters and subjects [CODESPLIT] mean.params <- t(apply(sampled$samples$alpha[,,keep],1:2,mean)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot the calibration curve (Abs vs. [BSA]) with an appropriate trendline and figure caption [CODESPLIT] ggplot(bca, aes(x = BSA, y = Abs562)) + ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "generate MPT data: [CODESPLIT] gendat <- genMPT(theta = pnorm(theta), restrictions = \"model/restrictions.txt\", numItems = numItems, eqnfile=\"model/2htsm.eqn\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Recode confirm(oppose/confirm) and won(win/lost) to factors [CODESPLIT] DF$confirm <- recode(DF$confirm, `0` = \"Confirmed\", `1` = \"Opposed\") DF$won <- recode(DF$won, `0` = \"Lost\", `1` = \"Won\") DF <- DF %>% mutate(confirm = as_factor(confirm)) %>% mutate(treatment = as_factor(treatment)) %>% mutate(won = as_factor(won)) %>% mutate(majmin = as_factor(majmin)) %>% mutate(experiment = as_factor(experiment)) %>% mutate(evidence = as_factor(evidence)) summary(DF) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create unique id number for each unique individual [CODESPLIT] mutate(uniq.id = group_indices(., id, experiment)) %>% mutate(uniq.id = as_factor(uniq.id)) summary(Ind_E1) M_Ind_E1 <- brm(changed ~ won*confirm + trial + (1|uniq.id), data = Ind_E1, family = \"bernoulli\", iter = 6000, chains = 3, cores = 3, save_all_pars = TRUE, file = \"Ind_E1\") summary(M_Ind_E1) plot(M_Ind_E1, ask = FALSE) pp_check(M_Ind_E1, check = \"distributions\") pp_check(M_Ind_E1, check = \"residuals\") pp_check(M_Ind_E1, \"error_scatter_avg\") pp_check(M_Ind_E1, check = \"scatter\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "generates 95% confidence intervals for each beta coefficient [CODESPLIT] m.p1_decision.CI <- round(confint(m.p1_decision, parm = \"beta_\"), 3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "exclude people who don't meet the inclusion criteria of having data for at least 3 days in each week [CODESPLIT] person_weekly_entries <- dat2019 %>% group_by(identity_id, woy)  %>% summarise(n_days = length(identity_id)) person_weekly_entries <- as.data.frame(person_weekly_entries) person_weekly_entries <- subset(person_weekly_entries, n_days >= 3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal glmer() model  This model contains a fixed withinsubjects effect of Ambiguity (effectcoded with 0.5 amb),  and betweensubjects fixed effects for Vocabulary and ART test scores,  plus random effects by participants and items.  [CODESPLIT] Acc.max <- glmer(accuracy ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CohOnly, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal lmer() model  This model contains a fixed withinsubjects effect of Ambiguity (effectcoded with 0.5 amb),  and betweensubjects fixed effects for Vocabulary and ART test scores,  plus random effects by participants and items.  Maximal model with interactions  [CODESPLIT] TrialDwellTime.max <- lmer(logTrialDwellTime ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " _firstfixation duration   Construct a maximal lmer() model  Because not all the necessary models converge, remove correlations between  random effects for all models for the firstfixation duration measure:  [CODESPLIT] AOIKey.FF.reduced <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 | RECORDING_SESSION_LABEL) + (0 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 | item) + (0 + Vocab.Cent | item) + (0 + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) AOICohcue.FF.reduced <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 | RECORDING_SESSION_LABEL) + (0 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 | item) + (0 + Vocab.Cent | item) + (0 + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "_gaze duration   Construct a maximal lmer() model [CODESPLIT] AOIKey.GazeDur.max <- lmer(logGazeDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) AOICohcue.GazeDur.max <- lmer(logGazeDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) AOISpillover.GazeDur.max <- lmer(logGazeDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "_regressions out   Construct a maximal glmer() model [CODESPLIT] AOIKey.RegrOut.max <- glmer(IA_REGRESSION_OUT ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) AOICohcue.RegrOut.max <- glmer(IA_REGRESSION_OUT ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) AOISpillover.RegrOut.max <- glmer(IA_REGRESSION_OUT ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) AOISpillover.RegrOut.max <- glmer(IA_REGRESSION_OUT ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "_firstfixation duration   Construct a maximal lmer() model [CODESPLIT] AOISpillover.FF.max <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) AOIWrapUp.FF.max <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) AOICohcue.FF.max <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent  | item), data = Data.CorrTrials, REML=FALSE) AOISpillover.FF.max <- lmer(logFirstFixDur ~ 1 + Ambiguity.code  + Vocab.Cent + ART.Cent + Ambiguity.code : Vocab.Cent + Ambiguity.code : ART.Cent + (1 + Ambiguity.code | RECORDING_SESSION_LABEL) + (1 + Vocab.Cent + ART.Cent | item), data = Data.CorrTrials, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Creating subsets of coins by time period and obtaining measures for each of the time periods [CODESPLIT] for (i in dates) { dsub <- (subset(df, df$DATE == i)) motifs <- cbind(dsub[318:681]) denom <- cbind(dsub[256:309]) DATE[i] <- i HDenomination[i] <- entropy(denom) HMotifs[i] <- entropy(motifs) CEDenominationsMotifs[i] <- condentropy(denom, motifs) NormCEDenominationsMotifs[i] <- condentropy(denom, motifs) / entropy(denom) CEMotifsDenominations[i] <- condentropy(motifs, denom) NormCEMotifsDenominations[i] <- condentropy(motifs, denom) / entropy(motifs) MI[i] <- mutinformation(denom, motifs) NDenominations[i] <- length(unique(dsub$DENOMINATION)) NCoins[i] <- length(unique(dsub$ID)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "STATISTICAL TESTING: nonparametric Spearman's correlation nonnormalized conditional entropy [CODESPLIT] cor.test(results$DATE, results$CEDenominationsMotifs, method = \"spearman\") rdates <- rev(dates) #dates = years BCE plot(results$DATE, results$CEDenominationsMotifs, xlim = c(600,330), xaxt='n', xlab = \"YEAR BCE\", ylab = \"H(D|d)\", main = \"P2: Conditional entropy of denomination given designs\") axis(1, at = rdates, labels = rdates) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots with H(D|d)  nonnormalized conditional entropy Getting mean, median and standard deviation across different authorities per period for nonnormalized conditional entropy of denomination given designs  [CODESPLIT] N <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = length) MEAN <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = mean) MEDIAN <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = median) SD <- aggregate(CEDenominationsMotifs ~ DATE, data = resultspoleis, FUN = sd) resultspoleis_summary <- cbind.data.frame(N, MEAN$CEDenominationsMotifs, MEDIAN$CEDenominationsMotifs, SD$CEDenominationsMotifs) colnames(resultspoleis_summary) <- c(\"DATE\",\"N\",\"MEAN\",\"MEDIAN\",\"SD\") resultspoleis_summary$SE <- resultspoleis_summary$SD / sqrt(resultspoleis_summary$N) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot mean and median H(D|d) across authorities per period [CODESPLIT] require(ggplot2) ggmean <- ggplot(resultspoleis_summary,aes(x=DATE,y=MEAN)) + labs(title =  \"P2: Mean conditional entropy of denominations given designs across authorities\", x = \"Year BCE\", y = \"mean H(D|d) across authorities\") + scale_x_reverse() + geom_errorbar(aes(ymin=resultspoleis_summary$MEAN-resultspoleis_summary$SE, ymax=resultspoleis_summary$MEAN+resultspoleis_summary$SE),width=.1) + geom_line() + geom_point() ggmean ggmedian <- ggplot(resultspoleis_summary,aes(x=DATE,y=MEDIAN)) + labs(title =  \"P2: Median conditional entropy of denominations given designs across authorities\", x = \"Year BCE\", y = \"median H(D|d) across authorities\") + scale_x_reverse() + geom_line() + geom_point() ggmedian ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "REGRESSION ANALYSIS: GROUPING BY AUTHORITIES  nonnormalized CE [CODESPLIT] resultspoleis$AUTHORITIES <- rownames(resultspoleis) resultspoleis$DATE <- as.numeric(as.character(resultspoleis$DATE))", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Meanimputes missing values for a vector x   NOTE: see http://www.mailarchive.com/rhelp@rproject.org/msg58289.html   Args:    x: numeric vector   Returns:    x, with missing values replaced by mean(x)  [CODESPLIT] return(replace(x, is.na(x), mean(x, na.rm = T))) } makeConstructMatrix <- function(net, a) { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "how many missing data for WVSES questions? [CODESPLIT] sum(is.na(df21$WVSE1)) sum(is.na(df21$WVSE2))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "To log transform Years [CODESPLIT] TEST_data <- TEST_data %>% mutate(Yearslog = log(Years))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Making factors for reasons based on factor analysis. This is making a new data frames that also has plans [CODESPLIT] TEST_data_withplans <- dplyr::select( TEST_data, c( 'Hierarchy':'Cooperate', 'plan_materials', 'plan_MS', 'self_MS_prestige', 'self_MS_cooper', 'self_materials_prestige', 'self_materials_cooper' ) ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Word Cloud GG plot [CODESPLIT] ggwordcloudc <- function(x,y){ freqt<-count(x,y) set.seed(42) ggplot(freqt, aes( label = y, size = n, color=factor(sample.int(10,nrow(x),replace=TRUE)))) + geom_text_wordcloud(area_corr = TRUE) + scale_size_area(max_size = 24) + theme_minimal() } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate quantiles for eigenvalues, but only store those from simulated CF model in percentile1  [CODESPLIT] percentile <- apply(parallel$values, 2, function(x) quantile(x, .95)) min <- as.numeric(nrow(obs)) min <- (4 * min) - (min - 1) max <- as.numeric(nrow(obs)) max <- 4 * max percentile1 <- percentile[min:max] percentile <- apply(parallel$values, 2, function(x) quantile(x, .95)) min <- as.numeric(nrow(obs)) min <- (4 * min) - (min - 1) max <- as.numeric(nrow(obs)) max <- 4 * max percentile1 = percentile[min:max] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Label the yaxis 'Eigenvalue' [CODESPLIT] scale_y_continuous(name = 'Eigenvalue') +", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Label the xaxis 'Factor Number', and ensure that it ranges from 1max  of factors, increasing by one with each 'tick' mark.  [CODESPLIT] scale_x_continuous(name = 'Factor Number', breaks = min(eigendat$num):max(eigendat$num)) + ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "drop row without fid data and without distance burrow data [CODESPLIT] fid <- fid[!is.na(fid$FID),] fid <- fid[!is.na(fid$dist_burrow),] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get probability contours for plot [CODESPLIT] kd <- ks::kde(plot_data, compute.cont=TRUE) contour_90 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"10%\"])[[1]]) contour_90 <- data.frame(contour_90) contour_80 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"20%\"])[[1]]) contour_80 <- data.frame(contour_80) contour_70 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"30%\"])[[1]]) contour_70 <- data.frame(contour_70) contour_60 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"40%\"])[[1]]) contour_60 <- data.frame(contour_60) contour_50 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"50%\"])[[1]]) contour_50 <- data.frame(contour_50) contour_40 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"60%\"])[[1]]) contour_40 <- data.frame(contour_40) contour_30 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"70%\"])[[1]]) contour_30 <- data.frame(contour_30) contour_20 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"80%\"])[[1]]) contour_20 <- data.frame(contour_20) contour_10 <- with(kd, contourLines(x=eval.points[[1]], y=eval.points[[2]], z=estimate, levels=cont[\"90%\"])[[1]]) contour_10 <- data.frame(contour_10) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "pcurve pcurve(m.cor) Funnel plot, symmetry test (Egger's regression), failsafe n  [CODESPLIT] pdf(\"Self_control_Funnel.pdf\",width=5,height=5) funnel(m.cor,xlab = \"Correlation\") dev.off() eggers.test(x = m.cor) fsn(yi=m.cor$TE, sei=m.cor$seTE, type=\"Rosenthal\") plot(copas(m.cor)) copas(m.cor) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " H3a: stress ~ trust  set Cauchy prior (0, 1) as stated in the preregistration  [CODESPLIT] prior.coef <- brms::prior(cauchy(0,1),class='b') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit logistic model to each individual's data to estimate PSEs [CODESPLIT] jdat$PSE <- 500 for(i in as.numeric(levels(jdat$Subject))){ dat <- subset(jdat, Subject==i) if(dat$cannot_fit[1]==0){ fit.glm <- glm(faster ~ Distort, family=binomial, data=dat) jdat$PSE[jdat$Subject==i] <- -coef(fit.glm)[1]/coef(fit.glm)[2]  #PSE is -b0/b1 } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effects model for graded judgments [CODESPLIT] mod_full <- lmer(GradedJudge ~ Context*(GMSI_Gen_Z + Order*Distort) + (1|Subject), data=jdat2) summary(mod_full) Anova(mod_full, type=3, test='Chisq') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Custom function for weighted Likert plot [CODESPLIT] panel <- function(...){ HH::panel.likert(...) vals <- list(...) df <- data.frame(x = vals$x, y = vals$y, groups = vals$groups) grps <- as.character(df$groups) for(i in 1:length(colnames(likertscipop.df))){ grps <- sub(paste0('^', colnames(likertscipop.df)[i]), i, grps) } df <- df[order(df$y,grps),] df$correctX <- ave(df$x, df$y, FUN = function(x){ x[x < 0] <- rev(cumsum(rev(x[x < 0]))) - x[x < 0]/2 x[x > 0] <- cumsum(x[x > 0]) - x[x > 0]/2 return(x) }) subs <- sub(' Positive$', '', df$groups) collapse <- subs[-1] == subs[-length(subs)] & df$y[-1] == df$y[-length(df$y)] df$abs <- abs(df$x) df$abs[c(collapse, F)] <- df$abs[c(collapse, F)] + df$abs[c(F, collapse)] df$correctX[c(collapse, F)] <- 0 df <- df[c(T, !collapse),] df$perc <- round(ave(df$abs, df$y, FUN = function(x){x/sum(x) * 100}), 1) df$perc <- paste0(df$perc,'%') df$perc[df$perc == \"0%\"] <- \"\" lattice::panel.text(x = df$correctX, y = df$y, label = df$perc, cex = 1.2, font = 1, col = \"white\") } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate mean and sd in deduplicated data set [CODESPLIT] means <- c(means, paste0(Round(mean(desc[!duplicated(desc$id), i])), \" (\", Round(sd(desc[!duplicated(desc$id), i])), \")\")) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " For each person combine component scores F, with loading matrix B and error values E to obtain latent structure  For each person standardized the data using the withinperson mean and standard deviation  [CODESPLIT] Y = matrix(0,nrow(data),P) colnames(Y) = sprintf(\"Y%d\",seq(1:P)) Error.var = matrix(0,nrow(data),P) colnames(Error.var) = sprintf(\"E%d\",seq(1:P)) FB.var = matrix(0,nrow(data),P) colnames(FB.var) = sprintf(\"FB%d\",seq(1:P)) for (i in 1:N){ n.i = which(data$ID==i) E = sqrt(var.E)*mvrnorm(length(n.i), rep(0,P), diag(P)) Y[n.i,] = F[n.i,] %*% t(B) + E Error.var[n.i,] = E FB.var[n.i,] = F[n.i,] %*% t(B) } data = cbind(data,Y) return(list(Sigma.Psi=NULL,Psi=Psi,Psi.i=NULL,lambda.max.list=NULL,Components=F,data=data, Error.var=Error.var,FB.var=FB.var)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Statistical analysis  Make GramGender a factor with two levels (feminine, masculine)  [CODESPLIT] subset.for.plot$GramGender <- as.factor(subset.for.plot$GramGender)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Summary statistics on the reported frequence of the structures [CODESPLIT] data_struc_names <- as.data.frame(sort(table(data$structures_names))) data_struc_names range(table(data$structures_names)) mean(table(data$structures_names)) sd(table(data$structures_names)) median(table(data$structures_names)) sort(table(data$structures_names))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exclude subjects who have missing values on the ability test or on selfviewed ability [CODESPLIT] mst2 <- subset(mst2, (Raven_self != \"NA\") & (Raven_obj != \"NA\") ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  DESCRIPTIVE STATISTICS   compute and save sample statistics (age distribution, number of females)  [CODESPLIT] age <- round(select(psych::describe(mst2_descr$age), n, min, max, mean, sd),2) age$n <- nrow(mst2_descr) sampstats <- mutate(age, female=plyr::count(mst2_descr$sex)[plyr::count(mst2_descr$sex)[,1]==\"1\",][\"freq\"] ) write.table(sampstats, file=\"Descriptives/age_sex_Sample_A_mst2.dat\", sep=\"\\t\", row.names=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get estimated variance of tanh1 (p hat) [CODESPLIT] dvartanh <- (1-(pe^2))^2 vartanh <- v/dvartanh vartanh", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We have weak evidence between precohort and coronacohort groups  test for strong invariance  fit model as multigroup model [CODESPLIT] est_s <- cfa(mod_s,quop_use, estimator = \"MLR\", missing = \"FIML\", group = \"corona\", group.equal = c(\"loadings\",\"intercepts\"), cluster = \"class\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "find the stations that have 90% of data after 2001 [CODESPLIT] Pdata_2000 <- Pdata[Dates >= \"2000-01-01\",] result <- rep(0,(ncol(Pdata_2000)-3)) for (i in 4:ncol(Pdata_2000)) { result[i-3] <- sum(ifelse(is.na(Pdata_2000[,i]),1,0))/nrow(Pdata_2000) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "result indicates the fraction of NA data for the stations  throw out all the columns and rows where result >0.1 [CODESPLIT] Pdata_new <- Pdata_2000[,-(which(result>0.1)+3)] Stations <- Stations[-which(result>0.1),] sub_b_sp <- SpatialPoints(cbind(Subbasins$Long_, Subbasins$Lat), proj4string=CRS(\"+proj=longlat +datum=WGS84 +no_defs\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " create dichotomous variable coded as 0  no, 1  yes  [CODESPLIT] dat[, pb_investigated_dich := ifelse(pb_investigated == \"yes\", 1, 0)] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " show ttest, cohen's d, and corrected pvals + CIs for all tests  [CODESPLIT] idx = 1 for (i in all_groups) { cat(\"\\n\") print(sprintf('TEST: %s, %s = %d', equation, grouping, i)) t <- t.test(formula = eval(parse(text = equation)), data = filter(df, .data[[grouping]] == i), paired = TRUE, conf.level = ci_level) print(t) print(sprintf('adjusted pval: %.4f', pvals_adj[idx])) d <- effsize::cohen.d(formula = eval(parse(text = paste0(equation, '| Subject(id)'))), data = filter(df, .data[[grouping]] == i), paired = TRUE) print(d) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " show wilcox test, cliff's d, and corrected pvals for all tests  [CODESPLIT] idx = 1 for (i in all_groups) { print(sprintf('TEST: %s, %s = %d', equation, grouping, i)) w <- wilcox.test(formula = eval(parse(text = equation)), data = filter(df, .data[[grouping]] == i), paired = TRUE) print(w) print(sprintf('adjusted pval: %.4f', pvals_adj[idx])) d <- effsize::cliff.delta(formula = eval(parse(text = equation)), data = filter(df, .data[[grouping]] == i), paired = TRUE) print(d) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit tree with quadratic model [CODESPLIT] hrstree <- semtree(fitted_quadratic_lgcm, as.data.frame(rndhrs_subset), control=semtree.control(verbose=TRUE, method=\"naive\", missing=\"party\", min.bucket = 500, min.N = 250, exclude.heywood = FALSE ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "then determine which response was given  unbiased guessing probablity of 1/N [CODESPLIT] p.guess <- rep(1/n.acc, n.acc) out$R[timer.replace] <- sample(1:n.acc, size=sum(timer.replace), replace=TRUE, prob=p.guess) } out } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " BassAckward EFA  We use bassAckward() to get the correlations between factors from successive solutions, so that we can create the hierarchical diagrams. The correlations are contained in the \"bass.ack\" result (See capture.output below).  In multifactor solutions, the factors obtained from fa() and bassAckward() can be matched by their sequential order (i.e., first factor in fa() is the same as the first factor from bassAckward()).  Note: BassAckwards is consistent with faCor but not factor.scores  [CODESPLIT] bass.usm=bassAckward(usm, nfactors=7, fm='ml', cut = .45, lr=F, items=F, plot=T) bass.uss=bassAckward(uss, nfactors=8, fm='ml', cut = .45, lr=F, items=F, plot=T) bass.its=bassAckward(its, nfactors=9, fm='ml', cut = .45, lr=F, items=F, plot=T) bassAckward.diagram(bass.its, lr=T, items=F, cut=.6) capture.output(bass.usm[[\"bass.ack\"]],file='bassAck USM.csv') capture.output(bass.uss[[\"bass.ack\"]],file='bassAck USS.csv') capture.output(bass.its[[\"bass.ack\"]],file='bassAck ITS.csv') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " In the loop below, users define the datasets and the number of solutions to obtain, and the script returns the fa() results, the structure matrices with item labels, and factor congruence coefficients across samples.  res stores complete fa results  load stores structure matrices  pv stores percent of variance accounted by the factors in each solution  fcong stores matrices with factor congruence coefficients  [CODESPLIT] for (i in 1:9) { # i = number of solutions we want for(s in seq_along(dl)) { # for each sample in \"dl\"... id=c('M','U','I')[s] # letter identifier for each sample: M = Mturk, U = US students, I = Italian students fa1=fa(dl[[s]],nfactors=i,fm='ml') # conduct EFA (oblimin rotation, ML estimation) pv[[s]][[i]]=fa1$Vaccounted[2,] # proportion variance accounted for by each factor colnames(fa1$Structure)=sub('ML',id,colnames(fa1$Structure)) # rename the factor labels according to sample res[[s]][[i]]=fa1 # append fa results to list load[[s]][[i]]=merge(unclass(fa.sort(fa1$Structure)), labels, by='row.names', sort=F) # attach item labels to structure loadings and append to list write.table(rbind(load[[s]][[i]],''), file=paste0(\"load\",s,\".csv\"), row.names=F, sep=',', append=T) # write matrix of structure loadings to file } fcong[[paste0(i,'_USM_USS')]]=fa.congruence(res[[1]][[i]],res[[2]][[i]],structure=T)  # append factor congruence coefficients for usm and uss fcong[[paste0(i,'_USM_ITS')]]=fa.congruence(res[[1]][[i]],res[[3]][[i]],structure=T) fcong[[paste0(i,'_USS_ITS')]]=fa.congruence(res[[2]][[i]],res[[3]][[i]],structure=T) } capture.output(fcong,file='fcong.csv') capture.output(pv,file='pvaccounted.csv') rm(fa1,fcong,load,pv,res,i,s,id) # cleaning afterwards ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Onefactor EFA    Here, we obtain the correlations among the four factors (obliquely rotated) and use them to find the factor loadings on a general factor  We also compute and save the factor scores for the 4 factors and the general factor  4factor EFA  [CODESPLIT] fa4.usm=fa(usm[,(items),with=F], nfactors=4, fm='ml') fa4.uss=fa(uss[,(items),with=F], nfactors=4, fm='ml') fa4.its=fa(its[,(items),with=F], nfactors=4, fm='ml') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1factor EFA using the factor scores (we need these scores in order to obtain the general factor scores) [CODESPLIT] fa1.usm=fa(usm[,(fnames),with=F], nfactors=1, fm='ml') fa1.uss=fa(uss[,(fnames),with=F], nfactors=1, fm='ml') fa1.its=fa(its[,(fnames),with=F], nfactors=1, fm='ml') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " in case xlim is set to 0, and some value has been given to xrange, center plot symmetrically on zero, using maximal extension in case range is set to 0, and given range otherwise  [CODESPLIT] if (xlim[1]==0 & !is.null(xrange))  { if (xrange==0)  { maxext <- max(abs(min(paramSampleVec)), abs(max(paramSampleVec)))  #largest extension into positive or negative range xlim = c(-maxext, maxext)   #centers plot symmetrically on zero } else { xlim = c(-xrange, xrange) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " compute and save correlation table of selfrated and objective ability measures and outcome aggregates,  [CODESPLIT] varnames <- c(\"Raven\",\"reasoning_self\",\"MWTB\",\"vocabulary_self\",\"global_selfevaluation\",\"well_being\",\"agency_self\",\"communion_self\",\"agency_peer\",\"communion_peer\") outcomes_pils <- dplyr::select(pils, Z_Raven_obj, Z_Raven_self, Z_MWTB_obj, Z_MWTB_self, Z_global_selfeval, Z_well_being, Z_agency_self, Z_comm_self, Z_agency_peer, Z_comm_peer) names(outcomes_pils) <- varnames cor_aggr <- corcons(outcomes_pils) write.table(cor_aggr, file=\"Descriptives/correlations_aggr_Sample_B_pils.dat\", sep=\"\\t\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function 5 Stouffer test for a vector of ppvalues [CODESPLIT] stouffer=function(pp) sum(qnorm(pp),na.rm=TRUE)/sqrt(sum(!is.na(pp))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1.4 Create family to turn t>F and z>chi2 [CODESPLIT] family=test family=ifelse(test==\"t\",\"f\",family) family=ifelse(test==\"z\",\"c\",family) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1.9 Take value after equal sign, the value of the teststatistic, and put it in vector \"equal\"  [CODESPLIT] equal=abs(as.numeric(substring(raw,eq+1)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute ppvalues for the half [CODESPLIT] pp33.half=ifelse(family==\"f\" & p<.025, (1/prop25)*(    pf(value,df1=df1,df2=df2,ncp=ncp33)-(1-prop25)),NA) pp33.half=ifelse(family==\"c\" & p<.025, (1/prop25)*(pchisq(value,df=df1,         ncp=ncp33)-(1-prop25)),pp33.half) pp33.half=pbound(pp33.half) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove lower triangle of correlation matrix [CODESPLIT] else if(removeTriangle[1]==\"lower\"){ Rnew <- as.matrix(Rnew) Rnew[lower.tri(Rnew, diag = TRUE)] <- \"\" Rnew <- as.data.frame(Rnew) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "mean RTs by condition [CODESPLIT] pretrain_study_rt <- pretrain_study %>% filter(block == 3) %>% group_by(id, condition) %>% summarise(mean_rt = mean(rt)) %>% ungroup() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "show get descriptive stats (how many/what proportion of participants reach each accuracy level) [CODESPLIT] pretrain_test_acc %>% group_by(test_rep) %>% summarise(n_80 = sum(mean_acc > 0.8), n_100 = sum(mean_acc == 1), prop_100 = sum(mean_acc == 1) / length(mean_acc), prop_80 = sum(mean_acc > 0.8) / length(mean_acc)) pretrain_test_acc %>% filter(test_rep == max(pretrain_test_acc$test_rep)) %>% summarise(group_mean_acc = mean(mean_acc), group_sd_acc = sd(mean_acc)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "P1 COUNTED as 6 seconds until sound event to normalize across trials P2 COUNTED AS FROM SOUND ONSENT (6SECONDS) plus 9 SECONDS  15 SECONDS P3 COUNTED AS 20 SECONDS FROM P2 TO ALMOST END OF THE TRIAL  ORDER PAPAMETER  'OPA', FISHER Z TRANSFORMED ORDER PARAMETER  'FOPA' P2 processing  [CODESPLIT] OPAmedianP2<-read.csv(file=\"OPAmedianP2.csv\", sep = \"\", header=FALSE) colnames(OPAmedianP2)<-c(colheaderChronos) DW_OPAmedianP2=add_column(OPAmedianP2, GR, .before = 1) FOPAmedianP2=FisherZ(OPAmedianP2) DW_FOPAmedianP2=add_column(FOPAmedianP2, GR, .before = 1) DL_OPAmedianP2=melt(DW_OPAmedianP2, id=c(\"GR\"), measured=c(\"colheaderChronos\")) DL_FOPAmedianP2=melt(DW_FOPAmedianP2, id=c(\"GR\"), measured=c(\"colheaderChronos\")) colnames(DL_OPAmedianP2)<-c(\"GrNr\", \"TrialNr\",\"OPAmedianP2\") colnames(DL_FOPAmedianP2)<-c(\"GrNr\", \"TrialNr\",\"FOPAmedianP2\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plotting correlations with qgraph [CODESPLIT] cor_graph <- qgraph(correlations)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "estimate regularized logistic nodewise regression network define where to binarize variables eLASSO (LASSO with EBIC model selection) listwise deletion of missing values (pairwise not possible for regressions) [CODESPLIT] Ising_net <- estimateNetwork(data, default = \"IsingFit\", split = split, missing = \"listwise\", rule = \"OR\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "standardize the variables (to obtain standardized coefficients) [CODESPLIT] dfAD[,c(\"S\",\"R\",\"H\")] <- data.frame(apply(dfAD[,c(\"S\",\"R\",\"H\")], 2, scale)) dfBE[,c(\"S\",\"R\",\"H\")] <- data.frame(apply(dfBE[,c(\"S\",\"R\",\"H\")], 2, scale)) dfCF[,c(\"S\",\"R\",\"H\")] <- data.frame(apply(dfCF[,c(\"S\",\"R\",\"H\")], 2, scale)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " run logistic regression on each subset, and predict the probability of  outcome for each country at each wave. these data points are used for the  metaanalysis  [CODESPLIT] response <- deparse(form[[2]]) data_list <- lapply(1:nrow(sub_data), function(i) { dat <- sub_data$data[[i]] if (!all_na(dat$migrant) && !all_na(dat[[response]]) && nlevels(droplevels(dat$migrant)) > 1) { m <- glm(form, family = binomial(), data = dat) pr <- ggemmeans(m, \"migrant\") pr$wave <- sub_data$wave[i] pr$country <- sub_data$country[i] pr } else { NA } }) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Binomial test for significant improvement in accuracy. [CODESPLIT] ifelse(rf_probs > .5, 0, 1) %>% sum %>% # count number of continuous responses binom.test(., nrow(pv), p = baseline) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Multiple item parameters and thresholds were in one cell. Split them into separate  columns [CODESPLIT] item.dif <- cbind(item.dif, str_split_fixed(item.dif$orig.parameter, \" \", 3)) item.dif <- cbind(item.dif, str_split_fixed(item.dif$orig.threshold, \" \", 2)) item.dif2 <- cbind(item.dif2, str_split_fixed(item.dif2$orig.parameter, \" \", 3)) item.dif2$Tau3 <- ifelse(item.dif2$item == \"CR021Q08\", 0.74200, NA) item.dif2 <- cbind(item.dif2, str_split_fixed(item.dif2$orig.threshold, \" \", 3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "interrater reliability create dataframe of reliability statistic for each of the 10 pairs of data [CODESPLIT] interraterReliability <- data.frame(row.names = row.names(interraterData)) for (i in c(1:10)) { interraterReliability[i,1] <- cohen.kappa(cbind(t(interraterData[i ,c(1:55)]), t(interraterData[i ,c(56:110)])))$weighted.kappa } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fill empty matrix with the counts significant (p < .05) per lag (rows) and per individual (column) [CODESPLIT] for (j in unique(dataset_imp$Participant)){ tempacf <- acf(dataset_imp[dataset_imp$Participant == j , -c(1,8:13)], lag.max = 60) acfmatrix[,paste0(\"participant\", j)] <- 0 for(i in as.numeric(rownames(acfmatrix))){ tempmatrix <- tempacf$acf[i,,] tempvector <- tempmatrix[upper.tri(tempmatrix, diag = T)] acfmatrix[i,paste0(\"participant\", j)] <- length(which(tempvector > 0.25 | tempvector < -0.25)) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create list with partial correlation matrices per window [CODESPLIT] pcorlist <- list() for (d in seq_len(30)) { pcorlist[[d]] <-  pcor(temp_data[temp_data$Time %in% d:(d+Window),c(\"Happy\", \"Worrying\",\"Nervous\",\"Act_later_regret\",\"Act_without_thinking\",\"Restless\")])$estimate } pcorlist <- rapply(pcorlist,function(x) ifelse(x==0.00000000,0.00000001,x), how = \"replace\") # will be pruned ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Smooth with (bayesian) logistic regression [CODESPLIT] (mym<-mean(nus)) d<-list(nus=nus-mym,fur=prob.fur*10000) m<-quap(alist( fur~dbinom(10000,p), logit(p)<-a+b*nus, a~dnorm(0,1), b~dnorm(0,1) ),data=d) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " For home (emotional) climate items, recode values of \"99\" to missing.  [CODESPLIT] mutate_at(vars(contains(\"EDNh_Emotion\")), ~ifelse(. == 99, NA, .)) %>%  # (Not sure exactly how this code works, but it does) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "figure < annotate_figure(figure, top  text_grob(str_c(Q_text[1]), size  8, color\"black\")) figure  [CODESPLIT] ggsave(str_c(ordinal_y[QN],\"_random.pdf\"),plot=figure,width=12,height=14) return(figure) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ordered_logistic stacked plot [CODESPLIT] N_cat <- max(d[,var_y],na.rm=T) grn <- 250 zLogM <- seq(-5,4,length.out=grn) P <- matrix(NA, nrow=grn, ncol=N_cat) for(i in 1:grn){ P[i,] <- ordered_logistic(fixef['b_X']*zLogM[i],cutpoints) } if(QN==1 | QN==4 | QN==5 | QN==8){ axis_labels <- c('fraction of responses','M-ratio') }else{ axis_labels <- c(' ',' ') } data.frame(P) -> P colnames(P) <- x_lab #str_c(\"p_\",1:N_cat) P$zlogM <- zLogM P$logM <- P$zlogM * 2*sd(log(d$M_cov)) + mean(log(d$M_cov)) P$M <- exp(P$logM) P %>% pivot_longer(cols=x_lab, #starts_with(\"p_\"), values_to=\"p\", names_to=\"k\") %>% mutate(k = as_factor(k), k = fct_rev(k)) %>% ggplot(aes(x=M, y=p, fill=k)) + scale_fill_viridis_d(name=\"\",labels = wrap_format(20), option=\"rocket\") + geom_area(alpha=0.8 , size=0.4, colour=\"black\")+ themeXstack + labs(y=axis_labels[1],x=axis_labels[2])+ scale_y_continuous(breaks=seq(0,1,0.1))+ coord_cartesian(xlim=c(0.5,1.3),ylim=c(0+0.045,1-0.045))+ ggtitle(label=str_wrap(sel_Q_text,22)) -> pl_stack ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate threshold for lowest quartile [CODESPLIT] quantile(tau_squared_self)[2]] tau2_thres_publ <- dat[as_factor(primary_data) == \"yes\" & k_publ > k_thres, quantile(tau_squared_self_publ)[2]] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "now we add up all corresponding answers to create one row per participant, as in a typical ANOVA design analysis  [CODESPLIT] sdt.agg <- sdt.agg %>% group_by(Participant, type) %>% summarise(count = n()) %>% spread(type, count) sdt.agg #this is a new dataframe for the aggregate data, you can delete after. ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " function pcor2beta gives you data from a partial correlation/network  input   pcor  a partial correlation matrix / network   output   a matrix of betas, each column corresponds to a dependent variable  so that you can get predicted values by a matrix multiplication  in the form betas %*% data  [CODESPLIT] pcor2beta <- function(pcor) { require(psych) require(corpcor) diag(pcor) <- 1 p <- ncol(pcor) betas <- matrix(0, ncol = p, nrow = p) for(i in 1:p) betas[-i,i] <- matReg(y = i, x = seq(p)[-i], C = pcor2cor(pcor))$beta betas[abs(betas) < 1e-13] <- 0 betas } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "extract posterior samples of omega [CODESPLIT] posterior.omega <- c(posterior.omega.check[, 1], posterior.omega.check[, 2]) posterior.omega.description.only <- c(posterior.omega.check[, 1], posterior.omega.check[, 2]) posterior.omega.description.plus.stats <- c(posterior.omega.check[, 1], posterior.omega.check[, 2]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "MEAN & SD FOR AMOUNT OF REM SLEEP (min) BETWEEN BASELINE AND ISOLATION [CODESPLIT] aggregate(REM~Place, data=REM_GNS, FUN=mean) aggregate(REM~Place, data=REM_GNS, FUN=sd) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculating sociality bias ratio variable  Converting the amount of dream interactions to same scale  as the amount of wake time interactions was reported [CODESPLIT] IN$int_cat<-ifelse(IN$Interactions == 0, 1, ifelse(IN$Interactions >= 1 & IN$Interactions <= 5, 2, ifelse(IN$Interactions >= 6 & IN$Interactions <= 15, 3, ifelse(IN$Interactions >= 16 & IN$Interactions <= 25, 4,5)))) IN$dream_per_sos<-IN$int_cat/IN$sos_int_prev_day hist(IN$dream_per_sos) qqnorm(IN$dream_per_sos) #skewed hist(log(IN$dream_per_sos)) qqnorm(log(IN$dream_per_sos)) #looks more normally distributed after log-transformed ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "join predicted to test data [CODESPLIT] test_pred <- left_join(test,prediction,by=\"id\") test_pred <- as.data.frame(test_pred) test_pred$class_var <- as.factor(test_pred$class_var) test_pred$class_var_pred <- as.factor(test_pred$class_var_pred) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Computing SMD (g) and its variance [CODESPLIT] smd <- escalc(measure = \"SMD\", m1i = wm, m2i = mm, sd1i = wsd, sd2i = msd, n1i = wn, n2i = mn, data = smd.means, append = TRUE) smd #Two columns (yi and vi) have been added# ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Computing Fisher's z and its variance [CODESPLIT] zcor <- escalc(measure = \"ZCOR\", ri = cor, ni = sample, data = zcor.correlations, append = TRUE) zcor ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Obtaining the forest plot [CODESPLIT] forest(res)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Analog to ANOVA: If we want to know the mean ES for each 'random' level: yes/no [CODESPLIT] res_r <- rma(yi = g, vi = var, mods = ~ factor(random)-1, data = dat) res_r ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Obtaining a funnel plot (model without moderators) [CODESPLIT] funnel(res, main = \"Random-Effects Model\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " inspect potential multicollinearity using the variance inflation factor (VIF),  for the example of the criterion variable outcome_sqd  (see Fox, 2016 for a discussion of VIFs and their cutoffs)  [CODESPLIT] lm_sqd <- lm(outcome_sqd ~ X + Y + X2 + XY + Y2, data=df) vif(lm_sqd) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mixed effects logistic regression [CODESPLIT] logistic_model_1 <- glmer(instructions ~ distance + (1|ID) + (1|target) + (1|task), data = choices, family = binomial(link = \"logit\")) logistic_model_2 <- glmer(instructions ~ distance + (1 + distance|ID) + (1|target) + (1|task), data = choices, family = binomial(link = \"logit\")) logistic_model_3 <- glmer(instructions ~ distance + (1 + distance|ID) + (1|target) + (1 + distance|task), data = choices, family = binomial(link = \"logit\"), control = glmerControl(optimizer = \"bobyqa\", optCtrl = list(maxfun = 100000))) model_comparison <- anova(logistic_model_1, logistic_model_2, logistic_model_3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "GLMMs  results of models with year as continuous are in table 3 Model with year as continuous variable [CODESPLIT] mod1ADI<-lmer(ADImS~LAT+LONG+YEAR1+(1|STATE)+(1|ROUTE2)+(1|SITE)+(1|YEAR1),dat=datUS,REML=F) mod1ADI<-lmer(ADImS~LAT+LONG+YEAR1+(1|SITE)+(1|COUNTRY)+(1|YEAR1),dat=datEU,REML=F) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "familiarity x female x BL   check effect of Baseline on familiarity x gender (self) interaction [CODESPLIT] HR_test_BL = lmer(HR_EMA ~ SI_familiarity_cw * female * HR_BL_cb + female * SI_gender_partner + SIAS_cb + State_SI_Anxiety_cw + SI_count + SI_type_simple + SI_duration_cw + SI_caffeine + SI_nicotin + SI_alcohol + accel_EMA_cw + (1|Participant), data=df_interact) summary(HR_test_BL) Anova(HR_test_BL, type=3) anova(HR_A, HR_test_BL) # no difference RMSSD_test_BL = lmer(ln_RMSSD_EMA ~ SI_familiarity_cw * female * ln_RMSSD_BL_cb + female * SI_gender_partner + SIAS_cb + State_SI_Anxiety_cw + SI_count + SI_type_simple + SI_duration_cw + SI_caffeine + SI_nicotin + SI_alcohol + accel_EMA_cw + (1|Participant), data=df_interact) summary(RMSSD_test_BL) Anova(RMSSD_test_BL, type=3) anova(RMSSD_A, RMSSD_test_BL) # better ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "let's try to create a boxplot for mpg on the  of cylinders (cyl) [CODESPLIT] ggplot(mtcars, aes(factor(cyl), mpg)) + geom_boxplot() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Likelihood Ratio Tests for nested model comparison [CODESPLIT] pair1_rand_anova <- anova(pair1_rand_full, pair1_rand2, pair1_rand3) # empty model = pair1_rand2 pair1_rand_anova_tidy <- tidy(pair1_rand_anova) pair2_rand_anova <- anova(pair2_rand_full, pair2_rand2, pair2_rand3) pair2_rand_anova_tidy <- tidy(pair2_rand_anova)  # create tidy table of model parameters pair2_fixed_anova <- anova(pair2_full, pair2_fixed, pair2_fixed2, pair2_fixed3, pair2_rand2) pair2_fixed_anova_tidy <- tidy(pair2_fixed_anova)  # create tidy table of model parameters pair2_fixed_lmer <- tidy(pair2_fixed2)  # create tidy table of model parameters group1_rand_anova <- anova(group1_rand_full, group1_rand2, group1_rand3, group1_rand4) # a random intercept for each pair is significantly better than empty, otherwise all other random effects tested were not warranted group1_rand_anova_tidy <- tidy(group1_rand_anova)  # create tidy table of model parameters group1_rand_lmer <- tidy(group1_rand3)  # create tidy table of model parameters group2_rand_anova <- anova(group2_rand_full, group2_rand2, group2_rand3, group2_rand4, group2_rand5) group2_rand_anova_tidy <- tidy(group2_rand_anova)  # create tidy table of model parameters group2_rand_lmer <- tidy(group2_rand2)  # create tidy table of model parameters ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fills in missing values of x with the mean of x   Args:    x: a numeric vector   Returns:    x with missing values filled in [CODESPLIT] return(replace(x, is.na(x), mean(x, na.rm = T))) } mseOfMatchingColumns <- function(nm, mat, dt) { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " **** 0.3.4) logisticPseudoR2s   calculates Logistic pseudo Rs (from Field et al., 2013)  input: glm object  [CODESPLIT] logisticPseudoR2s <- function(LogModel) { dev <- LogModel$deviance nullDev <- LogModel$null.deviance modelN <- length(LogModel$fitted.values) R.l <- 1 - dev / nullDev R.cs <- 1- exp ( -(nullDev - dev) / modelN) R.n <- R.cs / ( 1 - ( exp (-(nullDev / modelN)))) outdat <- data.table(R.l = R.l, R.cs = R.cs, R.n = R.n) outdat[, r.sq.print := paste(round(R.l, 3), \"/\", round(R.cs, 3), \"/\", round(R.n, 3), sep = \"\")] return(outdat) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "for loop goes through all the files in myLists and applies addTagsTextEntry to each row in the file [CODESPLIT] for (k in 1:length(myLists)) { i <- paste(\"list\", k, \".txt\", sep = \"\") o <- paste(\"upload\", k, \".txt\", sep = \"\") f <- read.delim(i, header = FALSE, stringsAsFactors = FALSE) write(\"[[AdvancedFormat]]\", o) apply(f, 1, addMultiChoiceSame, o) } rm(list=ls()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " compute mean LT per condition, group and accuracy  [CODESPLIT] LT <- ddply(data, .(rctype,group,acc_lab), summarize, mean.rt = mean(rt), se.lower = mean.rt - se(rt), se.upper = mean.rt + se(rt)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit uni.cfa and save output in uni.cfi.fit. Use fixed factor (std.lv  TRUE), and FIML for missing data)  [CODESPLIT] uni.cfa.fit = cfa(uni.cfa, data = SSSSdat, std.lv = TRUE, missing = \"ML\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Request summary output from model, including fit indexes, standardized estimates, and R^2/communalities  [CODESPLIT] summary(uni.cfa.fit, fit.measures = TRUE, standardized = TRUE, rsquare=TRUE) summary(five.cfa.fit, fit.measures = TRUE, standardized = TRUE, rsquare=TRUE) summary(five.sem.parcel.corr.fit, fit.measures = TRUE, standardized = TRUE, rsquare=TRUE) summary(five.sem.parcel.reg.fit, fit.measures = TRUE, standardized = TRUE, rsquare=TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We need to get an idea of how many factors are likely needed. Parallel analysis can help. Save output in object \"parallel\", using psych() fa.parallel function. Variables 125 (the BFI items) are to be analzyed, using maximum likelihood (ml) common factors (fa). Simulate 50 other samples of \"garbage factors\", using R^2s (SMC) as initial communality estimates and compare observed eigenvalues to 95th quantile of simulated \"garbage factor\" eigenvalues  [CODESPLIT] parallel = fa.parallel(SSSSdat[1:25], fm = 'ml', fa = 'fa', n.iter = 50, SMC = TRUE, quant = .95) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit measurement invariance models based on five.cfa model, distinguishing by group levels of \"gender\". Save output in invar.output object  [CODESPLIT] invar.output = measurementInvariance(five.cfa, data = SSSSdat, group = \"gender\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create function to calculate ICC from fitted model: ICC  between family variance / total family variance [CODESPLIT] calc.icc <- function(y) { sumy <- summary(y) (sumy$varcor$famnumber[1]) / (sumy$varcor$famnumber[1] + sumy$sigma^2) } calc.icc <- function(y) { sumy <- summary(y) (sumy$varcor$famnumber[1]) / (sumy$varcor$famnumber[1] + sumy$sigma^2) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "** 5.2) Poisson regression   **** 5.2.1) Full set   model 1: summary effect (absolute value) [CODESPLIT] pred.nbias.1 <- glm(nmeth_bias ~ scale(abs(MA_ES_self)), data = dat[!is.na(nmeth_bias) & k_self > k_thres, ], family = poisson) summary(pred.nbias.1) rsq.pred.nbias.1 <- logisticPseudoR2s(pred.nbias.1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " ANOVA dependent variable: drift rate independent variable:masking time condition  letters task, pretest  [CODESPLIT] cv_pre_drift = cogito %>% gather(key = \"maskingtime\", value = \"drift\", cvt1_v1,cvt1_v2,cvt1_v3,cvt1_v4) %>% convert_as_factor(obs, maskingtime) aov_cv_pre <- anova_test(data = cv_pre_drift , dv = drift, wid = obs, within = maskingtime, type=3, detailed=T) get_anova_table(aov_cv_pre, correction=\"none\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "subsample of the PG without patients with schizophrenia or other psychotic disorders (F20. or F23.) [CODESPLIT] subsample_PGwithoutF20_23 <- subsample_PG[subsample_PG$F20_23==0, ] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " function returns raster of posterior probabilities for bivariate normal data  x is the unknown tissue of interest, will have two values, one for each isotope  m is a 2D vector, all the values in the raster for each isotope  v is the same as m, but for variances  r is a single number  the covariance. Can be vector if estimated as nonstationary  ras is a raster that will serve as a template for the final product  [CODESPLIT] calcCellProb2D <- function(x,m,v,r,ras) { pd <- 1/(2*pi*sqrt(v[,1])*sqrt(v[,2])*sqrt(1-r^2))*exp(-(1/(2*(1-r^2)))* ((x[1]-m[,1])^2/v[,1]+(x[2]-m[,2])^2/v[,2]-(2*r*(x[1]-m[,1])* (x[2]-m[,2]))/(sqrt(v[,1])*sqrt(v[,2])))) pdras <- setValues(ras,pd) return(pdras) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function returns raster of posterior probability distribution [CODESPLIT] calcPostProb <- function(x){ pp <- x/cellStats(x,sum) return(pp) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "reproject the site coordinates to get the site to plot (thanks to P. Schauer) [CODESPLIT] myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords myproj <- proj4string(origins)#get the CRS string iso.data.sp <- iso.data #make a copy coordinates(iso.data.sp) <- ~ Long + Lat #specify the coordinates proj4string(iso.data.sp) <- CRS(\"+proj=longlat +datum=WGS84\") #project first iso.data.sp <- spTransform(iso.data.sp, myproj) #then re-project using the projection from the base map wrld_simpl2 <- wrld_simpl #make a copy wrld_simpl2 <- spTransform(wrld_simpl2, myproj) #this is to add in country borders etc. if you want to as they have the same projection issues as the site coords ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "cutting dendrogram in 3 clusters [CODESPLIT] clus3 = cutree(hcOSrenamel, 3) Women_WOCP_cluster<-plot(as.phylo(hcOSrenamel), type = \"fan\", tip.color = cbbPalette[clus3]) Women_WOCP_cluster ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate intercorrelations between diamonds on the betweenperson level with 95% Bootstrapped CI [CODESPLIT] between.person.diamonds = diamonds %>% group_by(user_id) %>% summarise(across(where(is.numeric), ~ mean(.x))) between.person.diamonds = between.person.diamonds %>% dplyr::select(diamonds_duty, diamonds_intellect, diamonds_adversity, diamonds_mating, diamonds_positivity, diamonds_negativity, diamonds_deception, diamonds_sociality) cors.diamonds = round(cor(between.person.diamonds, method = \"pearson\"),2) cors.diamonds = as.data.frame(cors.diamonds) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Analysis  Linear Probability Model on Impact Factor [CODESPLIT] fit_ev_if <- lm(X1.year.Impact.Factor ~ diff_ev + factor(year), data=FullDataset) summary(fit_ev_if) # show results fit_ex_if <- lm(X1.year.Impact.Factor ~ diff_ex + factor(year), data=FullDataset) summary(fit_ex_if) # show results ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Tobit Model as robusntess check  assign value 0 to Impact factors for unpublished studies [CODESPLIT] FullDataset$X1.year.Impact.Factor[FullDataset$published == \"No\"] = 0 tobit_ev_if <- censReg( X1.year.Impact.Factor ~ diff_ev + factor(year), data = FullDataset, left = 0) summary(tobit_ev_if) tobit_ex_if <- censReg( X1.year.Impact.Factor ~ diff_ex + factor(year), data = FullDataset, left = 0) summary(tobit_ex_if) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "initialize standard error function [CODESPLIT] stderr <- function(x, na.rm=FALSE) { if (na.rm) x <- na.omit(x) sqrt(var(x)/length(x)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "2019 data  zstandardize mood scores for each person [CODESPLIT] dat2019_scaled <- dat2019_complete %>% dplyr::select(identity_id, country, doy, mood, weekday) %>% group_by(identity_id)  %>% mutate_at(vars(-identity_id,-doy,-country, -weekday), scale) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Run regression anlayses and save as table (make one table with mood and depression  see depression section) [CODESPLIT] m1_het <- lm(mood_change ~ gender + age + education_level + race, data = dat_het_prepost) m2_het <- lm(depression_change ~ gender + age + education_level + race, data = dat_het_prepost) tab_model(m1_het, m2_het, show.est = F, show.std = T, digits = 3, file=\"Mood_Depression_PrePost_Demog_std.doc\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Run multilevel model analyses predicting mood scores from the the dummycoded month variable [CODESPLIT] dat2020_complete$month <- as.factor(dat2020_complete$month) summary(m1 <- lmer(depression ~ month + (1|identity_id), data = subset(dat2020_complete, country == \"United States\"))) summary(m2 <- lmer(depression ~ month + (1|identity_id), data = subset(dat2020_complete, country == \"Germany\"))) summary(m3 <- lmer(depression ~ month + (1|identity_id), data = subset(dat2020_complete, country == \"United Kingdom\"))) tab_model(m1,m2,m3, show.est = F, show.std = T, digits = 3, file=\"Depression_2020_Months.doc\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Step 2: Create onehot encodings (dummy variables) [CODESPLIT] testData3 <- predict(dummies_model, testData2) testData3 <- predict(dummies_model, testData2) testData3 <- predict(dummies_model, testData2) testData3 <- predict(dummies_model, testData2) testData3 <- predict(dummies_model, testData2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " function to generate output csv files for a list of response IDs  inputs: dataset, list of response IDs  outputs: csv files named with response IDs  [CODESPLIT] multi_response <- function(dataset, IDs_list) { for (i in IDs_list) { participant_response <- single_response(dataset, i) csv_name <- str_c(i, \".csv\", sep = \"\") write.csv(participant_response, csv_name) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Means and SDs of proportional looking time to novel object in the preferential looking phase  condition 1 [CODESPLIT] mean(dfX_base$PrefLook_LT_Object_Nov_PROP[which(dfX_base$Condition == \"Con1\")], na.rm = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate mode of a vector [CODESPLIT] mode.knn =  function(x){ uniq.x = unique(x) uniq.x = uniq.x[which(!is.na(uniq.x))] knn = uniq.x[which.max(tabulate(match(x, uniq.x)))] return(knn) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "find [k] neigherst neigbors and impute by their mode [CODESPLIT] impute.knn = function(y, k){ t = which(is.na(y)) if(length(t) == 0){ return(y) }else{ is = 1:length(t) for(i in is){ if(i > k){ look.at = y[(t[i]-k):(t[i]+k)] y[t[i]] = mode.knn(look.at) } if(i <= k){ look.at = y[1:(t[i]+k)] y[t[i]] = mode.knn(look.at) } } return(y) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get categorical data information then look up the variables in the process  in categorical_data so we can get the number of categories for each variable in the process  this would be achieved by categories[cat_indices[i]] for variable i in the process [CODESPLIT] categories <- as.integer(mplus.get.group.attribute(file,'categorical_data','categories')) catvars <- mplus.get.group.attribute(file,'categorical_data','var_names') vartypes <- as.integer(mplus.get.group.attribute(file,'categorical_data','vtype')) if (series) { categories <- as.integer(mplus.get.group.attribute(file,'categorical_data','categories')) catvars <- mplus.get.group.attribute(file,'categorical_data','var_names') vartypes <- as.integer(mplus.get.group.attribute(file,'categorical_data','vtype')) if (series) { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get indices and names of the variables in the series [CODESPLIT] var_indices <- mplus.get.group.attribute(file,cstr2,'var_indices') var_names <- mplus.get.group.attribute(file,cstr2,'var_names') cat_indices <- pmatch(var_names, catvars, nomatch=0) cat_indices <- as.integer(cat_indices) var_indices <- mplus.get.group.attribute(file,cstr2,'var_indices') var_names <- mplus.get.group.attribute(file,cstr2,'var_names') cat_indices <- as.integer(pmatch(var_names, catvars, nomatch=0)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "extract correlations between selfreported scores (observational level!) [CODESPLIT] self.reported.scores = c(\"duty\", \"intellect\", \"mating\", \"positivity\", \"sociality\") l = \"mating\" for(i in self.reported.scores[1:length(self.reported.scores)-1]){ data_phi = ftable(data[,c(paste0(\"diamonds_\", l), paste0(\"diamonds_\", self.reported.scores[which(self.reported.scores == i)+1]))]) print(paste(l, \"and\", self.reported.scores[which(self.reported.scores == i)+1])) print(phi(data_phi)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "generate unique study identifier, format and annotate data  [CODESPLIT] d$id <- paste0(d$First_Author,\", \", d$Year) d$Year <- as.numeric(substr(d$Year,1,4)) d$ids <- NA for(i in d$id) d[d$id == i,\"ids\"] <- 1:dim(d[d$id == i,\"ids\"])[1] d$ids <- as.character(d$ids) d$Age <- (as.numeric(d$Age_M) - 30) / 20 d$energy_renew <- d$energy_renew1 + d$energy_renew2 find_cCode <- Vectorize(function(i) which(unlist(Map(function(x) sum(d[i,grepl(\"SVS\",names(d))] - f[x,grepl(\"SVS\",names(f))]), 1:dim(f)[1])) == 0)) d$Ccode <- f$Ccode[find_cCode(1:dim(d)[1])] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " to make sure that effects are tested from complex to simple, we reverse the order of the vector containing the to be tested effects:  [CODESPLIT] effects_to_test = rev(effects_to_test)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "combine model matrix and data to include participant id and DV: [CODESPLIT] new_data = data.frame(dplyr::select(data, contains(c(group, DV_variables, by))), model_matrix) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check if retention is a number from 0 to 1 [CODESPLIT] if (retention < 0 || retention > 1) { stop('Retention value is not a number from 0 to 1.') }", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal glmer() model  This model contains code for Ambiguity effect,  plus random effects by participants and items.  [CODESPLIT] Acc.Modality.max <- glmer(Correct ~ 1 + Ambiguity.code + (1 + Ambiguity.code | Ã¯..ID) + (1   | Item), data = Data.DisambTask.AmbvsUnamb, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a histogram of the residuals [CODESPLIT] hist(rawResiduals) hist(invResiduals) hist(logResiduals)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "construct a maximal lmer() model  This model contains a fixed withinsubjects effect of Ambiguity (effectcoded with 0.5 amb)  plus random effects by participants and items. [CODESPLIT] RT.max <- lmer(logRT ~ 1 + Ambiguity.code + (1 + Ambiguity.code | Ã¯..ID) + (1   | Item), data = Data.DisambTask.AmbvsUnamb, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal glmer() model  This model contains codes for Modality effects,  plus random effects by participants and items.  [CODESPLIT] Acc.Modality.max <- glmer(Correct ~ 1 + Modality.code1 + Modality.code2 + (1 + Modality.code1 + Modality.code2 | Ã¯..ID) + (1   | Item), data = Data.DisambTask.Amb, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.Modality.max <- glmer(True.Positive ~ 1 + Modality.code2 + (1 + Modality.code2 | ID) + (1   | Item), data = Data.ListAndRead, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.Modality.max <- glmer(True.Positive ~ 1 + Modality.code3 + (1 + Modality.code3 | ID) + (1   | Item), data = Data.ListAndRSVP, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.Modality.max <- glmer(True.Positive ~ 1 + Modality.code3 + (1 + Modality.code3 | ID) + (1   | Item), data = Data.ReadAndRSVP, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal glmer() model  This model contains codes for run effects,  plus random effects by participants and items.  [CODESPLIT] Acc.Run.max <- glmer(Correct ~ 1 + Run.code1 + Run.code2 + (1 + Run.code1 + Run.code2 | Ã¯..ID) + (1   | Item), data = Data.DisambTask.Amb, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.Run.max <- glmer(True.Positive ~ 1 + Run.code1 + Run.code2 + (1 + Run.code1 + Run.code2 | ID) + (1   | Item), data = Data.RecMem, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run a onesample ttest comparing accuracy to chance level (0.5) [CODESPLIT] t.test(Data.List$True.Positive, mu=0.5) t.test(Data.Read$True.Positive, mu=0.5) t.test(Data.RSVP$True.Positive, mu=0.5) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "abbreviate first names to match with abbreviations in publication list [CODESPLIT] stats_persons <- lapply(stats_persons, function(x){ x[2]<-str_sub(x[2], 1, 1) x }) stats_persons <- lapply(stats_persons, function(x)paste(x[1], x[2], sep = \",\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Load the data [CODESPLIT] load('../Relative Effectiveness Data - final.Rdata')", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Print a txt and csv file with the results ' ' @param stats_to_print the tibble/table to print ' @param name_file name of the file to write ' ' @return 2 files with the results (txt and csv) [CODESPLIT] print_result <- function(stats_to_print, name_file){ knitr::kable( stats_to_print, format = \"rst\")  %>% cat( file = here('results', str_c(name_file, '.txt', sep = \"\")), sep = \"\\n\") stats_to_print %>% write_csv(file = here('results', str_c(name_file, '.csv', sep = \"\"))) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "convert degrees of freedom into numeric variables and store in new variable [CODESPLIT] ScienceStatus$df.numerator.value = as.numeric(as.character(ScienceStatus$df.numerator)) ScienceStatus$df.denominator.value = as.numeric(as.character(ScienceStatus$df.denominator))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Assigning numeric values to variable labels which were stored in CSV [CODESPLIT] ScienceStatus$coding.difficulty_r = revalue(ScienceStatus$coding.difficulty, c(\"Very Easy\"=\"1\", \"Moderately Easy\"=\"2\", \"Slightly Easy\"=\"3\", \"Neither Difficult nor Easy\"=\"4\", \"Slightly Difficult\"=\"5\", \"Moderately Difficult\"=\"6\", \"Very Difficult\"=\"7\")) ScienceStatus$coding.difficulty_r = as.numeric(as.character(ScienceStatus$coding.difficulty_r)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calcuate EXACT pvalue from the stats reported in the paper (Posthoc) [CODESPLIT] ScienceStatus_J$Calc.Pvalue<-(rowSums(cbind(ScienceStatus_J$T.pvalue_calc, ScienceStatus_J$F.pvalue_calc, ScienceStatus_J$rg.pvalue_calc, ScienceStatus_J$r.pvalue_calc,ScienceStatus_J$chi.pvalue_calc), na.rm = TRUE) + ifelse(is.na(ScienceStatus_J$T.pvalue_calc) & is.na(ScienceStatus_J$F.pvalue_calc) & is.na(ScienceStatus_J$rg.pvalue_calc) & is.na(ScienceStatus_J$r.pvalue_calc) & is.na(ScienceStatus_J$chi.pvalue_calc), NA, 0)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "PCurve Graph      Calculated from statistic and DF   By Paper [CODESPLIT] P.By.Paper<-ddply(ScienceStatus_SP, .(article.id,yearcat), summarize, Pmean.calc = stouffer.P(Calc.Pvalue.Clean), PMedian.calc = median(Calc.Pvalue.Clean)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "    RIndex    Calcuate Mean, Median, Peak  [CODESPLIT] Rindex.Reults<-BCa.Boot.CI(ScienceStatus_SP,Calc.Z,r.index.calc.boot,LogT=FALSE,splitter=yearcat, StatType=\"AR\") Rindex.Reults ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "FIGURE 3 Length in phonemes    RD  fit a model with fixed effects and with the random effect of subjects on intercepts and the slopes for length [CODESPLIT] Cue.Subj.lmer = glmer(ACC ~ CueCondition + zLengthPh + zFreq + (1+zLengthPh|Subject), data = NamingData, family = \"binomial\", control = glmerControl(optimizer=\"bobyqa\")) summary(Cue.Subj.lmer) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "extract fitted values and align with original data for plotting [CODESPLIT] a<-fitted.values(Cue.Subj.lmer) length(a) Dat<-na.omit(NamingData) #remove NAs from original data so aligns with model values Dat$fitted<-a names(Dat) a<-fitted.values(Cue.Subj.lmer) length(a) Dat<-na.omit(NamingData) #remove NAs from original data so aligns with model values Dat$fitted<-a names(Dat)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "FIGURE 4a Plot average slope for effect of Frequency [CODESPLIT] p <- ggplot(data=Dat,aes(x=zFreq, y=fitted)) p <- p + geom_smooth(method = \"glm\") p <- p + xlab(\"Frequency (z score)\") + ylab(\"Accuracy (model fit)\") p <- p + ggtitle(\"4a) Average (group) effect of Frequency\") p <- p + ylim(0,1) p <- p + theme_bw() + theme(text=element_text(size=12)) p ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Main effects model for predicting inversion questions during alignment sessions. [CODESPLIT] inversion_main_effects_model<-glmer(inversion_dv ~ prime_type + WMC + modality +  inversion_production_pre + (1 + prime_type + trial_order | subject) + (1|verb), priming, family=\"binomial\"(link=\"logit\"),glmerControl(optimizer = \"bobyqa\", optCtrl = list(maxfun = 100000))) summary(inversion_main_effects_model) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Odds ratios for main effects model [CODESPLIT] inversion.me.CI<-confint(inversion_main_effects_model, parm=\"beta_\", level=0.90, method=\"Wald\") inversion.me.tab <- cbind(est = fixef(inversion_main_effects_model),inversion.me.CI) inversion.me.tab <- exp(inversion.me.tab) inversion.me.tab <- as.data.frame(inversion.me.tab) inversion.me.tab <- rownames_to_column(inversion.me.tab) inversion.me.tab ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Visualize the interaction between prime type and modality.  save the specific effect to a variable [CODESPLIT] effect2 <-effect(\"prime_type*modality\",inversion_full_model_sig_int_only) summary(effect2) plot(effect2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "post hoc analyses to compare all levels of test order  between groups indirect production main effect only [CODESPLIT] ind_prod_me_between_groups <- emmeans(indirect_production_me,c(\"test_order\"), type = \"response\") pairs(ind_prod_me_between_groups, reverse = F, type = 'response', adjust = 'none') plot(ind_prod_me_between_groups) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Test assumptions MANOVA  Test whether residuals are normally distributed [CODESPLIT] df$pc1.residuals = lm(pc1~condition.socaccount, data=df)$residuals df$pc2.residuals = lm(pc2~condition.socaccount, data=df)$residuals df$pc3.residuals = lm(pc3~condition.socaccount, data=df)$residuals df$pc4.residuals = lm(pc4~condition.socaccount, data=df)$residuals shapiro.test(df$pc1.residuals) shapiro.test(df$pc2.residuals) shapiro.test(df$pc3.residuals) shapiro.test(df$pc4.residuals) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Only include tweets with at least 3 words [CODESPLIT] final_twitter_data_ex <- final_twitter_data %>% filter(wc>=3) final_twitter_data_ex <- final_twitter_data_ex %>% filter(bryscore>=0.0001)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Means and SDs in feartweets and anxietytweets [CODESPLIT] final_twitter_data_ex %>% group_by(fearVSanx) %>% summarise_at(vars(bryscore,i,we,they,focusfuture), list(mean=mean, sd=sd)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Independent samples ttest: Brysbaert concreteness score as DV [CODESPLIT] t.test_bci <- final_twitter_data_ex %>% rstatix::t_test(bryscore ~ fearVSanx, var.equal = TRUE, detailed=TRUE) %>% rstatix::add_significance() t.test_bci ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "convert age from months to years [CODESPLIT] demog.tab$interview_age <- demog.tab$interview_age / 12", "target": 1, "target_options": ["no_match", "match"]}
{"input": "multinomial CI (these are simulataneous CI although assume full pooling) [CODESPLIT] obs_p_CI <- DescTools::MultinomCI(pl_d$N, conf.level=1-0.05/2, method=\"goodman\") *4 pl_d$obs_p_lb <- obs_p_CI[,2] pl_d$obs_p_ub <- obs_p_CI[,3] pl_d %>% mutate(resp_true = ifelse(k>=4,1,0), fill_group = str_c(item, resp_true, type), line_group = str_c(item, type), item = factor(item, levels=c(\"true\",\"false\"))) %>% ggplot(aes(x=k,y=obs_p))+ facet_grid(item ~ type) + nice_theme + geom_col(aes(fill=fill_group))+ geom_errorbar(aes(y=obs_p,ymin=obs_p_lb, ymax=obs_p_ub),color=\"dark grey\",width=0,lwd=0.8)+ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mratio posterior density plot [CODESPLIT] log_mu_sci <- unlist(c(output[,grep('mu_logMratio_1',varnames(output))])) log_mu_cov <- unlist(c(output[,grep('mu_logMratio_2',varnames(output))])) d_groupM <- data.frame(logM =c(log_mu_sci,log_mu_cov), type=c(rep('science',length(log_mu_sci)),rep('Covid-19',length(log_mu_sci)))) d_groupM %>% mutate(M=exp(logM)) %>% ggplot(aes(x=M,y=type,fill=type))+ nice_theme + geom_vline(xintercept = 1, lty=2,size=0.4)+ stat_halfeye(.width = c(.95, .95),aes(fill=type),alpha=0.8)+ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert `campaign_week` from a character string to a factor with 11 levels to order weeks chronologically. [CODESPLIT] Facebook_Users$campaign_week <- factor(Facebook_Users$campaign_week, levels = c(\"Week 1\", \"Week 2\", \"Week 3\", \"Week 4\", \"Week 5\", \"Week 6\", \"Week 7\", \"Week 8\", \"Week 9\", \"Week 10\", \"Week 11\", \"Election Day\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert `candidate_page` from a character string to a factor with 3 levels.(Facilitates data visualization and data exploration). [CODESPLIT] Facebook_Users$candidate_page <- factor(Facebook_Users$candidate_page, levels = c (\"Harper\", \"Trudeau\", \"Mulcair\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Adds a new column to identify the Facebook page being `liked`. [CODESPLIT] Harper_User_Likes$partisanship <- \"Conservative\" Trudeau_User_Likes$partisanship <- \"Liberal\" Mulcair_User_Likes$partisanship <- \"Social Democrat\" ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Match partisan assignment from Facebook_User_Likes to the Facebook Users Dataset using `user_id`. [CODESPLIT] Facebook_Users$partisanship <- NA Facebook_Users$partisanship <- Facebook_User_Likes$partisanship[match(Facebook_Users$user_id, Facebook_User_Likes$user_id)] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert `partisanship` from a character string to a factor with 3 levels.(Facilitates data visualization and data exploration). [CODESPLIT] Facebook_Users$partisanship <- factor(Facebook_Users$partisanship, levels = c (\"Conservative\", \"Liberal\", \"Social Democrat\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "split the data and transform it into integers [CODESPLIT] ARTE <- as.numeric(data[which(data$TV.Station == 'ARTE'),]$Rating) Degeto <- as.numeric(data[which(data$TV.Station == 'Degeto Film'),]$Rating) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run the onesided ttest with alpha  0.1 [CODESPLIT] t.test(ARTE, Degeto, alternative = \"greater\", conf.level = 0.90) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create list of dataframes per country [CODESPLIT] countries <- unique(data8$cntry_full) index = 0 listofdfs <- list()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Draw correlation plots [CODESPLIT] char.col<-c(\"#228822\",\"#663388\",\"#006688\") groupn<-c(4,11,6,3) left<--8 top<-25 corrplot.mixed(cortabNB, upper = \"ellipse\",tl.pos=\"lt\",tl.col=rep(char.col,groupn[1:3]),number.cex = .7) text(left,top,\"Non-biological\\nfathers\",cex=1.2,pos=4) nc<-0.5 w<-16 tiff(\"Figure_S3_corr1.tif\",width=w,height=2*w,units=\"cm\",res=600,compression=\"lzw\") par(mfrow=c(2,1)) corrplot.mixed(cortabNB, upper = \"ellipse\",tl.pos=\"lt\",number.cex = nc,tl.col=rep(char.col,groupn[1:3])) text(left,top,\"Non-biological\\nfathers\",cex=1.2,pos=4) corrplot.mixed(cortabPA, upper = \"ellipse\",tl.pos=\"lt\",number.cex = nc,tl.col=rep(char.col,groupn[1:3])) text(left,top,\"Partners\",cex=1.2,pos=4) dev.off() tiff(\"Figure_S4_corr2.tif\",width=w,height=2*w,units=\"cm\",res=600,compression=\"lzw\") par(mfrow=c(2,1)) corrplot.mixed(cortabLO, upper = \"ellipse\",tl.pos=\"lt\",number.cex = nc,tl.col=rep(char.col,groupn[1:3])) text(left,top,\"Sensitive\\nperiod\\nlog-odds\",cex=1.2,pos=4) corrplot.mixed(cortabD, upper = \"ellipse\",tl.pos=\"lt\",number.cex = nc,tl.col=rep(char.col,groupn[1:3])) text(left,top,\"Relative\\nsimilarity\\nt value\",cex=1.2,pos=4) dev.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  function to calculate proporiton of studies within each time period for continuous variables that were transformed in a cutoff   x  character string of the variable  name  character string of the name that will appear in the table  time  character string of the time period variable  method  either \"chisq\" or \"fisher\" depending on the type of the test wanted  [CODESPLIT] CAT<-function(x, name, time, method) { tabtot <- table(DATA[,x]) sumtot <- sum(tabtot) tabtime <- table(DATA[,x], DATA[,time]) sumtime <- apply(tabtime, 2, sum) proptot <- tabtot[2]/sum(tabtot) CItot <- exactci(tabtot[2], sum(tabtot), conf.level = 0.95) proptime <- tabtime[2,]/sumtime CItime <- map2(tabtime[2,], sumtime, function(x, y){exactci(x, y, conf.level = 0.95)}) propCItime <- map2(proptime, CItime, function(x,y){paste0(round(x*100,0),\" (\", round(y[[\"conf.int\"]][1]*100,0),\"-\", round(y[[\"conf.int\"]][2]*100,0),\")\")}) if(method == \"fisher\"){ p <- fisher.test(tabtime) } else if(method == \"chisq\"){ p<-chisq.test(tabtime) } N_propCItime <- NULL for(i in 1:length(propCItime)){ N_propCItime <- c(N_propCItime, c(sumtime[i], propCItime[[i]])) } res=c(name, sumtot, paste0(round(proptot*100,0),\" (\",round(CItot$conf.int[1]*100,0),\"-\", round(CItot$conf.int[2]*100,0),\")\"), N_propCItime, round(p$p.value[1],4)) return(res) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to calculate overall effect size for death or readmission rates (output has results + model to do model checks)   var  character string of the variable to calculate the effect size for [CODESPLIT] rate_ES <- function(var){ dat <- na.omit(DATA[,c(\"number_follow_up\", var)]) colnames(dat)<-c(\"Ni\",\"Ei\") dat_es <- escalc( xi= Ei, ni =Ni, data = dat, measure = \"PLO\", to=\"if0all\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  function to get weighted logistic regression with continuous X variable and proportion Y variable and ggplot showing the predicted relationship   x  X variable as character  y  Y variable as character  Xpred  vector of X values to get predictions for  labX  X label as character  labY  Y label as character  limY  axis limits for Y axis, vector of 2 values  title  plot title as character  ypos  y relative position of text with OR and pvalue  [CODESPLIT] wtlogis_plot <- function(NB, x, y, Xpred, labX, labY, limY, title, ypos, unit){ dat <- data.frame(count=as.integer(DATA[,y]), Ntot=DATA[,NB], t=DATA[,x]) %>% mutate(prop = count/Ntot, perc = prop*100, wt = log(Ntot), t_div = t/10) dat <- na.omit(dat) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "model with time divided by 10 to get OR for 10yr increment [CODESPLIT] timediv_mod <- glm(prop ~ t_div, family = \"binomial\", weights = wt, data = dat) ORdiv <- tidy(timediv_mod, conf.int = TRUE) %>% mutate(OR_CI = paste0(round(exp(estimate),2), \" (\", round(exp(conf.low),2),\"-\", round(exp(conf.high),2), \")\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  function to get weighted linear regression with continuous X variable and continuous Y variable and ggplot showing the predicted relationship   x  X variable as character  y  Y variable as character  Xpred  vector of X values to get predictions for  labX  X label as character  labY  Y label as character  limY  axis limits for Y axis, vector of 2 values  title  plot title as character  ypos  y relative position of text with OR and pvalue  [CODESPLIT] wtlin_plot <- function(NB, x, y, Xpred, labX, labY, limY, title, ypos, unit){ dat <- data.frame(y=as.integer(DATA[,y]), Ntot=DATA[,NB], t=DATA[,x]) %>% mutate(wt = log(Ntot), t_div = t/10) dat <- na.omit(dat) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to get OR and CI from the model with 10 year increment   obj  model with x variable divided by 10  p  pvalue from the LRT [CODESPLIT] OR_div <- function(obj, p){ OR_CI <- paste0(round(exp(obj$beta[2]),2), \" (\", round(exp(obj$ci.lb[2]),2),\"-\", round(exp(obj$ci.ub[2]),2), \")\") pval_clean <- ifelse(p[[\"pval\"]] < 0.001, \"p<0.001\", paste0(\"p=\",round(p[[\"pval\"]],3))) OR_p <- c(OR_CI, p, pval_clean) return(OR_p) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  function to get metaregression model with continuous X variable and proportion Y variable and ggplot showing the predicted relationship   x  X variable as character  y  Y variable as character  Xpred  vector of X values to get predictions for  labX  X label as character  labY  Y label as character  limY  axis limits for Y axis, vector of 2 values  title  plot title as character  ypos  y relative position of text with OR and pvalue  [CODESPLIT] MA_prop_plot <- function(NB, x, y, Xpred, labX, labY, limY, title, ypos){ prop_dat <- na.omit(DATA[,c(NB, y, x)]) colnames(prop_dat)<-c(\"Ni\",\"Ei\",\"X\") prop_dat$prop <- (prop_dat$Ei/prop_dat$Ni) * 100 prop_dat$X_div <- prop_dat$X/10 es_prop_dat <- escalc( xi= Ei, ni =Ni, data = prop_dat, measure = \"PLO\", to=\"if0all\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to get OR for subgroup variables with 2 categories   x  character string for the x variable  y  character string for the y variable  group  chracter string for the group variable  newref  character string for the level of the group variable that is not the reference level [CODESPLIT] OR_group_fun <- function(x, y, group, newref){ OR_data <- data.frame(Group = rep(NA,2), N = rep(NA,2), OR = rep(NA,2), lowCI = rep(NA,2), upCI = rep(NA,2), p = rep(NA,2), I2 = rep(NA,2)) group_dat <- na.omit(DATA[,c(\"number_follow_up\", y, x, group)]) colnames(group_dat)<-c(\"Ni\",\"Ei\",\"X\",\"group\") group_dat$X_div <- group_dat$X/10 OR_data$Group <- levels(group_dat$group) OR_data$N <- as.integer(table(group_dat$group)) es_group_dat <- escalc( xi= Ei, ni =Ni, data = group_dat, measure = \"PLO\", to=\"if0all\") inter_mod1 <- rma(yi=yi,mods = ~ X_div * group, vi=vi, data=es_group_dat, method = \"REML\") inter_mod2 <- rma(yi=yi,mods = ~ X_div * I(relevel(group, newref)) , vi=vi, data=es_group_dat, method = \"REML\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  function to get metaregression model with continuous X1 and X2 variables and 2 group variables and proportion Y variable and ggplot showing the predicted relationship with x1 for a median value of x2   x1  X1 variable as character  x2  X2 variable as character  group1  group1 variable (2level factor) as character  group2  group2 variable (2level factor) as character  y  Y variable as character  x1pred  vector of values for which to predict x1  x2val  median value of X2 for which to get predictions  labX  X label as character  labY  Y label as character  limY  axis limits for Y axis, vector of 2 values  title  plot title as character  ypos  y relative for text of OR and pvalue  [CODESPLIT] MA_prop_plot_adj <- function(NB, x1, x2, group1, group2, y, x1pred, x2val, labX, labY, limY, title, ypos){ prop_dat <- na.omit(DATA[,c(NB, y, x1, x2, group1, group2)]) colnames(prop_dat)<-c(\"Ni\",\"Ei\",\"X1\", \"X2\", \"group1\", \"group2\") prop_dat$prop <- (prop_dat$Ei/prop_dat$Ni) * 100 prop_dat$X1_div <- prop_dat$X1/10 es_prop_dat <- escalc( xi= Ei, ni =Ni, data = prop_dat, measure = \"PLO\", to=\"if0all\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " rayleigh_test   Rayleigh test for the uniformity of directional data,  as described by Mardia et al. (1979: chapter 15) and Mardia & Jupp (1999)  Note that there would probably be better alternatives on CRAN  [CODESPLIT] rayleigh_test <- function(X, correction = TRUE, check = TRUE, convert = TRUE) { if(check) { if(!isTRUE(all.equal(diag(tcrossprod(X)), 1))) { if(convert) { X / sqrt(diag(tcrossprod(X))) warning(\"X was converted into directional data\") } else { stop(\"Directional data (in coordinates) are assumed as X\") } } } p <- ncol(X) n <- nrow(X) Mean <- colMeans(X) Norm_Mean_2 <- drop(crossprod(Mean)) S <- p * n * Norm_Mean_2 if(correction) { S <- S * (1 - 1 / 2 / n) + S ^ 2 / (2 * n * (p + 2)) } P_value <- pchisq(S, p, lower.tail = FALSE) list(Norm_Mean = sqrt(Norm_Mean_2), n = n, p = p, Statistics = S, P_value = P_value) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute ICC  run null/unconditional model and use intercept and residual variances to solve for ICC (i.e., Intercept Variance / (Intercept Variance + Residual Variance))  [CODESPLIT] Model.Null <- lmer(VerbPhysAggSum ~ 1 + (1 | ID), data = LineBisectionStudy) summary(Model.Null) ICC <- (18.69/(18.69+7.76)) print(ICC) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function to extract legend (from: https://gist.github.com/crsh/be88be19233f1df4542aca900501f0fbfilegglegendrL7)  otherwise plots without tracks would have no legend [CODESPLIT] gglegend <- function(x){ tmp <- ggplot_gtable(ggplot_build(x)) leg <- which(sapply(tmp$grobs, function(y) y$name) == \"guide-box\") tmp$grobs[[leg]] } legend = gglegend(gi) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "define function doing the bootstrap: [CODESPLIT] boot.fun<-function(x, xcall., data., rv.name., m., discard.non.conv., save.path., extract.all.){ xdone=F while(!xdone){ done2=F while(!done2){ data.[, rv.name.]=simulate(object=m.)[, 1] if(xcall[[\"xfam\"]][[\"family\"]]!=\"beta\" | (xcall[[\"xfam\"]][[\"family\"]]==\"beta\" & min(data.[, rv.name.])>0 & max(data.[, rv.name.])<1)){ done2=T } } i.res=try(update(m., data=data.), silent=T) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "prepare model frame for conditional model for prediction: [CODESPLIT] model=attr(terms(as.formula(xcall[[\"cond.form\"]])), \"term.labels\") model=model[!grepl(x=model, pattern=\"|\", fixed=T)] if(length(model)==0){model=\"1\"} cond.m.mat=model.matrix(object=as.formula(paste(c(\"~\", paste(model, collapse=\"+\")), collapse=\"\")), data=new.data) if(set.circ.var.to.zero){ cond.m.mat[,paste(c(\"sin(\", circ.var.name, \")\"), collapse=\"\")]=0 cond.m.mat[,paste(c(\"cos(\", circ.var.name, \")\"), collapse=\"\")]=0 } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function that draws any bellcurve in the form and shape as the ones used in the paper [CODESPLIT] docurve<-function( sd=2, dist=1, xcent=10, ycent=0.6, paroff=0.15, offoff=0.1, percoff=0.06, percoff2=0.03, thresh.vis=0.01, bellims=15, botlims=8, pardown=0.01, xoff=0, par.col=\"#0097BD\", off.col=\"#FF0066\", off.in.col=\"#FF0066\", off.out.col=\"#FF6600\", fading=\"33\", col.bot=\"#555555\", lwd.par=1.2, lwd.bot=1.2){ off.in.col2<-paste(off.in.col,fading,sep=\"\") off.out.col2<-paste(off.out.col,fading,sep=\"\") p1<-xcent-dist/2 p2<-xcent+dist/2 c1<-gen.curve(sd,average=0,bellims=bellims) polygon(replast(c1[[1]][c1[[1]]<(-dist/2)])+xcent,c(c1[[2]][c1[[1]]<(-dist/2)],0)+ycent,col=off.out.col2,border=NA) polygon(repfirst(c1[[1]][c1[[1]]>(dist/2)])+xcent,c(0,c1[[2]][c1[[1]]>(dist/2)])+ycent,col=off.out.col2,border=NA) polygon(repboth(c1[[1]][c1[[1]]>=(-dist/2)&c1[[1]]<=(dist/2)])+xcent,c(0,c1[[2]][c1[[1]]>=(-dist/2)&c1[[1]]<=(dist/2)],0)+ycent,col=off.in.col2,border=NA) lines(c(xcent-botlims,xcent+botlims),c(ycent,ycent),col=col.bot,lwd=lwd.bot) lines(c(p1,p1),c(ycent-pardown,ycent+paroff),col=par.col,lwd=lwd.par) lines(c(p2,p2),c(ycent-pardown,ycent+paroff),col=par.col,lwd=lwd.par) outprob<-sum(c1[[2]][c1[[1]]>dist/2])/sum(c1[[2]]) inprob<-sum(c1[[2]][c1[[1]]<=dist/2&c1[[1]]>=-dist/2])/sum(c1[[2]]) offtail<-mean(c(c1[[1]][c1[[1]]>dist/2][1], c1[[1]][c1[[1]]>dist/2][!(c1[[2]][c1[[1]]>dist/2]>thresh.vis)][1])) text(xcent,ycent+percoff,paste(round(inprob*100),\"%\",sep=\"\"),col=off.in.col,cex=0.8) text(c(xcent-offtail,xcent+offtail)+c(-1,1)*xoff,ycent+percoff2,paste(round(outprob*100),\"%\",sep=\"\"),col=off.out.col,cex=0.8) text(c(p1,p2),ycent+paroff,c(expression(t[p[1]]),expression(t[p[2]])),pos=3,col=par.col,offset=0.15) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "loglikelihood for each subject using their mean parameter vector [CODESPLIT] mean_pars_ll <- numeric(ncol(mean_pars)) data <- transform(sampled$data, subject = match(subject, unique(subject))) for (j in 1:nsubj) { mean_pars_ll[j] <- sampled$ll_func(mean_pars[j, ], data = data[data$subject == j,], sample = FALSE) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create the plots and populate them. The yaxt  \"n\" serves to remove axis labels from the yaxis while the axis function serves to add axis labels in line with APS standards for graph publication. The lines and points functions add the connecting lines and symbols to the graphs. We repeat the process three times to create the three graphs corresponding to different values of tau.  [CODESPLIT] plot(k, end.data.cov[ ,1], type = \"l\", ylab = \"Coverage\", main = \"A: Tau = 0.2\", yaxt = \"n\", ylim = c(0.9, 1),xlab = expression(\"Number of Studies \" (italic(k)))) axis(side = 2, at = c(0.92,0.94,0.96,0.98,1),labels = c(\".92\",\".94\",\".96\",\".98\",\"1.0\")) points(k, end.data.cov[ ,1], pch = 0) lines(k, end.data.cov[ ,2]) points(k, end.data.cov[ ,2], pch = 3) lines(k, end.data.cov[ ,7]) points(k, end.data.cov[ ,7], pch = 5) lines(k, end.data.cov[ ,14]) points(k, end.data.cov[ ,14], pch = 4) legend(\"bottomright\", c(\"bWT-DL\",\"bWT-REML\", \"bWT-HE\", \"Sub-Q\"), pch = c(0,5,4,3), cex = 0.9, ncol = 2) plot(k, end.data.cov[ ,3], type = \"l\", ylab = \"Coverage\", main = \"B: Tau = 0.4\", yaxt = \"n\", ylim = c(0.9, 1),xlab = expression(\"Number of Studies \" (italic(k)))) axis(side = 2, at = c(0.92,0.94,0.96,0.98,1),labels = c(\".92\",\".94\",\".96\",\".98\",\"1.0\")) points(k, end.data.cov[ ,3], pch = 0) lines(k, end.data.cov[ ,4]) points(k, end.data.cov[ ,4], pch = 3) lines(k, end.data.cov[ ,9]) points(k, end.data.cov[ ,9], pch = 5) lines(k, end.data.cov[ ,17]) points(k, end.data.cov[ ,17], pch = 4) legend(\"bottomright\", c(\"bWT-DL\",\"bWT-REML\", \"bWT-HE\", \"Sub-Q\"), pch = c(0,5,4,3), cex = 0.9, ncol = 2) plot(k, end.data.cov[ ,5], type = \"l\", ylab = \"Coverage\", main = \"C: Tau = 0.6\", yaxt = \"n\", ylim = c(0.9, 1),xlab = expression(\"Number of Studies \" (italic(k)))) axis(side = 2, at = c(0.92,0.94,0.96,0.98,1),labels = c(\".92\",\".94\",\".96\",\".98\",\"1.0\")) points(k, end.data.cov[ ,5], pch = 0) lines(k, end.data.cov[ ,6]) points(k, end.data.cov[ ,6], pch = 3) lines(k, end.data.cov[ ,11]) points(k, end.data.cov[ ,11], pch = 5) lines(k, end.data.cov[ ,20]) points(k, end.data.cov[ ,20], pch = 4) legend(\"topright\", c(\"bWT-DL\",\"bWT-REML\", \"bWT-HE\", \"Sub-Q\"), pch = c(0,5,4,3), cex = 0.9, ncol = 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Define the initial column name for x/y coordinates you want to use [CODESPLIT] xy_columns = list( x = \"GazePointX (ADCSpx)\", y = \"GazePointY (ADCSpx)\" ), ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compute cutoff for guessing by taking the 99% quantile from the binomial distribution (given by guessing probability and number of tests) [CODESPLIT] mutate(guessing_probability = 1/19, cut_off = qbinom(p = .99, size = number_tests, prob = guessing_probability)/number_tests, guessing_check = ifelse(mean_acc > cut_off, TRUE, FALSE)) %>% group_by(participant_id, condition) %>% summarize(guessing_check =  as.logical(min(guessing_check))) %>% ungroup() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "split the data into different data frames depending on the intended analysis   create dataframe with data on the hebb effect for the main analysis of the learning data [CODESPLIT] data_hebb_task = data_filtered %>% filter(phase == \"WM\") %>% select(participant_id, condition, phase, block, trial_number, hebb_trial, presentation_order:duration) %>% ungroup() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check cluster agreement using hierarchical clustering on raw items instead of factors [CODESPLIT] HCheck.HClusterD <- dist(EFA.mainData[, c(2:56)], method = \"euclidean\") HCheck.HClusterFit <- hclust(HCheck.HClusterD, method=\"ward.D2\") HCheck.HCluster <- cutree(HCheck.HClusterFit, k = 3) HCheckCluster <- data.frame(Factor = HCluster, Item = HCheck.HCluster) HCheckCluster <- cbind(HCheckCluster, ifelse(HCheckCluster[1] == HCheckCluster[2], 1, 0)) colnames(HCheckCluster)[3] <- 'Agreement' HCheckClusterAgreement <- HCheckCluster[, c(2:3)] %>% group_by(Item) %>% summarise(total = n(), n_agree = sum(Agreement))  %>% rename(Cluster = Item) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create dataframe with factor scores and cluster membership [CODESPLIT] KClusterScores <- cbind(EFA.mainData[, c(111:116)], KClusterFit$cluster) colnames(KClusterScores)[7] <- \"KCluster\" ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "testing homogeneity of variance in factors [CODESPLIT] KClusterScores %>% pivot_longer(c(Sensory, CognitiveDemand, ThreatToSelf, CrossSettings, Safety, States), names_to = c(\"Factor\")) %>% group_by(Factor) %>% levene_test(value ~ as.factor(KCluster)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "univariate Welch's ANOVA test with Bonferroni correction [CODESPLIT] table1 <- KClusterScores %>% pivot_longer(c(Sensory, CognitiveDemand, ThreatToSelf, CrossSettings, Safety, States), names_to = c(\"Factor\")) %>% mutate(Factor = factor(Factor, levels = c(\"Sensory\", \"CognitiveDemand\", \"ThreatToSelf\", \"CrossSettings\", \"Safety\", \"States\"))) %>% group_by(Factor) %>% welch_anova_test(value ~ as.factor(KCluster)) %>% adjust_pvalue(method = \"bonferroni\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "manually calculating omega squared and adjusted confidence intervals for each test [CODESPLIT] table1$omegasq <- apply(table1, 1, function(x) omega.F(as.numeric(x[5]), as.numeric(x[6]), as.numeric(x[4]), as.numeric(x[3]), 0.05/6)$omega) table1$omegalow <- apply(table1, 1, function(x) omega.F(as.numeric(x[5]), as.numeric(x[6]), as.numeric(x[4]), as.numeric(x[3]), 0.05/6)$omegalow) table1$omegahigh <- apply(table1, 1, function(x) omega.F(as.numeric(x[5]), as.numeric(x[6]), as.numeric(x[4]), as.numeric(x[3]), 0.05/6)$omegahigh) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effect model (accuracy) [CODESPLIT] LDTnonword_ACC_LME = glmer(accuracy ~ DISHARc + (1+DISHARc|item) + (1|subject), data = byTrial, family=binomial, control=glmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(LDTnonword_ACC_LME) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Estimate d from t and df [CODESPLIT] dest <-function (t,df) { r2 <- (t^2) / (t^2 + df) d.est <- (2*(r2^.5)) / (1-r2) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Stimuli categorized according to a multinomial distribution [CODESPLIT] y[i,] ~ dmulti(r[i,], t[i]) predy[i,1:ncat] ~ dmulti(r[i,], t[i]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Denominator is just the sum of the numerator values [CODESPLIT] denominator[i] = sum(numerator[i,]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "annotate lines so that each  has a unique id for ggplot overplotting  (else two lines from the same draw but  different replicates can get confused  with each other) [CODESPLIT] func_samples_surface <- func_samples_surface %>% mutate(line_id = as.numeric(rownames(func_samples_surface))) func_samples_aer <- func_samples_aer %>% mutate(line_id = as.numeric(rownames(func_samples_aer)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "cross product decay_rates with x (time) values  and calculate y (titer) values [CODESPLIT] cat('setting up x values...\\n') to_plot_surface <- func_samples_surface %>% crossing(surface_plot_times) to_plot_aer <- func_samples_aer %>% crossing(aer_plot_times) to_plot_surface <- to_plot_surface %>% mutate(predicted_titer = 10^(intercept - decay_rate * time)) to_plot_aer <- to_plot_aer %>% mutate(predicted_titer = convert_mL_media_to_L_air * 10^(intercept - decay_rate * time)) max_nonzero_time <- to_plot_surface %>% filter(log10(predicted_titer) > lowest_log_titer) %>% select(time) %>% max() surface_xlim <- c(0, max_nonzero_time) aer_xlim <- c(0, aer_max_x) print(aer_xlim) aer_jitwid <- 3/100 fit_panel_surface <- to_plot_surface %>% ggplot(aes( x = time, y = predicted_titer, color = virus, group = line_id)) + geom_line(alpha = line_alpha, size = line_size) + scale_colour_manual(values = unlist(virus_colors)) + geom_point(aes(x = time, y = 10^(log10_titer), group = trial_unique_id), data = surface_dat, color = pointborder, fill = pointfill, alpha = pointalpha, size = pointsize, stroke = pointstroke, position = position_jitter( width = jitwid, height = jith, seed = 5)) + geom_hline( data = experiment_dat_virus_surface, aes(yintercept = detection_limit), linetype = detection_linestyle, size = detection_linesize) + scale_y_continuous(trans = ytrans, breaks = ybreaks, labels = yformat) + coord_cartesian(ylim = surface_ylim, xlim = surface_xlim) + facet_grid(vars(virus), vars(material), drop = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " group adjustment for sigma with prior 0,0.1 for the beta  [CODESPLIT] ACT.BF.sigma.prior01 <- stan(data=ACT.data,file=\"./BayesFactor/models/ACT/ACT.BF.sigma.prior01.stan\", iter=40000, warmup =1000, chains=3,control = list(adapt_delta = 0.9)) ACT.BF.sigma.prior01.bridge <- bridge_sampler(ACT.BF.sigma.prior01) saveRDS(ACT.BF.sigma.prior01.bridge, \"./BayesFactor/marginal_lik/ACT/ACT.BF.sigma.prior01.rds\") DA.BF.sigma.prior01 <- stan(data=DA.standata,file=\"./BayesFactor/models/DA/DA.BF.sigma01.stan\", iter=40000, warmup =1000, chains=3) DA.BF.sigma.prior01.bridge <- bridge_sampler(DA.BF.sigma.prior01) saveRDS(DA.BF.sigma.prior01.bridge, \"./BayesFactor/marginal_lik/DA/DA.BF.sigma.prior01.rds\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "increase in rsq when using elastic net with items instead of linear regression with sum scores (without covariates) [CODESPLIT] tab_delta_rsq <- merge(tab[learner.id == \"elastic net\" & predictors == \"items\" & control == \"excluded\", .(elasticnet_items = rsq.test.mean), by = c(\"target\", \"design\")], tab[learner.id == \"linear regr\" & predictors == \"sum scores\" & control == \"excluded\", .(linearregr_sumscores = rsq.test.mean), by = c(\"target\", \"design\")]) tab_delta_rsq[, .(mean_delta_rsq = mean(elasticnet_items - linearregr_sumscores)), by = \"design\"] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate Cohen's d_z [CODESPLIT] return_Cohen_d_z <- function(variable_1, variable_2){ mean_differences <- mean(variable_1) - mean(variable_2) sd_var1 <- sd(variable_1) sd_var2 <- sd(variable_2) cor_var1_var2 <- cor(variable_1, variable_2) Cohen_d_z <- abs(mean_differences) / sqrt( (sd_var1 ^ 2) + (sd_var2 ^ 2) - (2 * sd_var1 * sd_var2 * cor_var1_var2) ) return(Cohen_d_z) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "visualizing the data add +scale_x_discrete(guide  guide_axis(angle  45)) to the plot to change the angle of the lables on the X axis [CODESPLIT] ggplot(mydata_p6_la, aes(x = Condition, y = Voltage)) + geom_boxplot() + facet_grid(.~ Specificity,)+ scale_x_discrete(guide = guide_axis(angle = 45)) ggplot(mydata_p6_ra, aes(x = Condition, y = Voltage)) + geom_boxplot() + facet_grid(.~ Specificity,)+ scale_x_discrete(guide = guide_axis(angle = 45)) ggplot(mydata_p6_lp, aes(x = Condition, y = Voltage)) + geom_boxplot() + facet_grid(.~ Specificity,)+ scale_x_discrete(guide = guide_axis(angle = 45)) ggplot(mydata_p6_rp, aes(x = Condition, y = Voltage)) + geom_boxplot() + facet_grid(.~ Specificity,)+ scale_x_discrete(guide = guide_axis(angle = 45)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fill output matrix with the calculated social measure [CODESPLIT] socialmeasures[,3] <- measure1 socialmeasures[,4] <- measure2 socialmeasures[,5] <- measure3 socialmeasures[,6] <- measure4 socialmeasures[,7] <- measure5 socialmeasures[,8] <- measure6 socialmeasures[,10] <- measure8 socialmeasures[,11] <- measure9 socialmeasures[,12] <- measure10 socialmeasures[,3] <- measure1 socialmeasures[,4] <- measure2 socialmeasures[,5] <- measure3 socialmeasures[,6] <- measure4 socialmeasures[,7] <- measure5 socialmeasures[,8] <- measure6 socialmeasures[,10] <- measure8 socialmeasures[,11] <- measure9 socialmeasures[,12] <- measure10 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "7 calculate the AVERAGE SHORTEST PATH of each node of the matrix [CODESPLIT] measure7 <- mean_distance(graphN, directed = TRUE, unconnected = TRUE) print(measure7) measure7 <- mean_distance(graphN, directed = TRUE, unconnected = TRUE) print(measure7) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Add Age Bins  This acts to help preserve privacy and allows easy plotting:     e.g., ggplot(Data, aes(x  Agebins) ) + geom_bar()  [CODESPLIT] Agebin = c('< 20', '20-29', '30-39','40-49','50-59','60-69','70-79','80+') Data %<>% mutate(Agebins = case_when( Age  < 20 ~ Agebin[1], Age >= 20 & Age < 30 ~ Agebin[2], Age >= 30 & Age < 40 ~ Agebin[3], Age >= 40 & Age < 50 ~ Agebin[4], Age >= 50 & Age < 60 ~ Agebin[5], Age >= 60 & Age < 70 ~ Agebin[6], Age >= 70 & Age < 80 ~ Agebin[7], Age >= 80  ~ Agebin[8] ) ) %>% apply_labels(Agebins = c('< 20' = 1, '20-29' = 2, '30-39' = 3, '40-49' = 4,'50-59' = 5,'60-69' = 6, '70-79' = 7,'80+' = 8)) Agebin = c('< 20', '20-29', '30-39','40-49','50-59','60-69','70-79','80+') Data %<>% mutate(Agebins = case_when( Age  < 20 ~ Agebin[1], Age >= 20 & Age < 30 ~ Agebin[2], Age >= 30 & Age < 40 ~ Agebin[3], Age >= 40 & Age < 50 ~ Agebin[4], Age >= 50 & Age < 60 ~ Agebin[5], Age >= 60 & Age < 70 ~ Agebin[6], Age >= 70 & Age < 80 ~ Agebin[7], Age >= 80  ~ Agebin[8] ) ) %>% apply_labels(Agebins = c('< 20' = 1, '20-29' = 2, '30-39' = 3, '40-49' = 4,'50-59' = 5,'60-69' = 6, '70-79' = 7,'80+' = 8)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ChiSquared  Gender create dataframe of gender and cluster membership, then exclude participant with 'Other' as response to analyse with Chisquared test  [CODESPLIT] data.OtherExcluded <- demographics %>% select(Gender, KCluster) %>% filter(Gender != 'Other') data.OtherExcluded$Gender <- droplevels(data.OtherExcluded$Gender) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "posthoc pairwise Chisquared with Bonferroni correction [CODESPLIT] pairwiseNominalIndependence(ASD.contable, compare = \"row\", fisher  = FALSE, gtest   = FALSE, chisq   = TRUE, method  = \"bonferroni\", digits  = 3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compile parameter estimates and fit of population ODE/SDE models [CODESPLIT] modt <- readRDS(\"PSM_transits_ODE.RDS\") mods <- readRDS(\"PSM_population_ODE.RDS\") i <- length(mods) pars <- c(round(modt[[4]]$THETA,1), \"OMEGA_stress\"=1.5, \"OMEGA_ke\"=.05, \"OMEGA_kt\"=.15, \"OMEGA_init\"=.15) pars[c(\"init\",\"sigma\")] <- c(.001,.001) parA <- list(LB=pars*.2, Init=pars, UB=pars*2.5) #bounds + inits npars <- Vectorize(function(x) sum(parA$Init != round(mods[[x]]$THETA,3)))(1:i) res <- Vectorize(function(x) round(mods[[x]]$THETA,3))(1:i) #parameter estimates for each fitted model res <- rbind(res, \"Mtt\"= Vectorize(function(x) round(4/mods[[x]]$THETA[\"kt\"],3))(1:i)) #add mean transit time for each model res <- rbind(res, \"LL\"= -Vectorize(function(x) mods[[x]]$NegLogL)(1:i)) res <- rbind(res, \"AIC\"= 2*npars +2*Vectorize(function(x) mods[[x]]$NegLogL)(1:i)) res <- rbind(res, \"R2\"= Vectorize(function(x) 1 - mods[[x]]$THETA[\"S\"]/var(PKdata$DV))(1:i)) colnames(res) <- letters[1:i] print(res) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " define a fixed grid of cutoff (threshold) values from [1, 0] with length `resolution`  [CODESPLIT] resolution <- 500 cutoff_out <- seq(1, 0, length.out = resolution) resolution <- 500 cutoff_out <- seq(1, 0, length.out = resolution) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Compute F1 skill score... ' ...for a given cutoffthreshold. ' @param sim vector of numeric values between [0, 1] (e.g., proportion of unstable grid points) ' @param obs vector of logicals (TRUE/FALSE) stating whether the layer was observed (of concern) or not. ' @param cutoff in percentage within (0, 1] ' @return numeric value of skill score ' @export  [CODESPLIT] calculateF1Score <- function(sim, obs, cutoff) { tp <- length(sim[sim >= cutoff & obs]) fn <- length(sim[sim < cutoff & obs]) fp <- length(sim[sim >= cutoff & !obs]) f1 <- 2*tp / (2*tp + fp + fn) return(f1) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Underscore split: split up the \"variable\" column in order to get the different factors (verbtype, input, moment)  [CODESPLIT] participantsdata.long <- cbind(participantsdata.long, colsplit(participantsdata.long$variable, \"_\", names = c(\"verbtype\", \"input\", \"testmoment\"))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "LDA  training the model [CODESPLIT] model_lda = train(class ~ ., data=trainSet, method='lda', trControl = trainControl(method = \"cv\")) fitted <- predict(model_lda) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " function to extract the sequence \"Obj_xxx\" untile a period \".\" appears  [CODESPLIT] slice_object_string <- function(string) { gsub(\".*Obj_(.*)\\\\..*\", \"\\\\1\", c(string)) } familiar_objects <- unlist(lapply(string_list, slice_object_string)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ChiSquare tests Relations between diet type and major themes [CODESPLIT] chisq.test(data$Diet, data$Health) #p = 0.94, chi-squared = 0.01 chisq.test(data$Diet, data$Food) #p = 0.44, chi-squared = 0.59 chisq.test(data$Diet, data$Social) #p = 0.80, chi-squared = 0.06 chisq.test(data$Diet, data$Logistic) #p = 0.17, chi-squared = 1.90 chisq.test(data$Diet, data$Finance) #p = 0.06, chi-squared = 3.61 chisq.test(data$Diet, data$Motivat.) #p = 0.76, chi-squared = 0.10 chisq.test(data$Diet, data$Diet.Cons.) #p = 0.05, chi-squared = 3.97 chisq.test(data$Diet, data$Other) #p = 1, chi-squared = 8.90e-29 chisq.test(data$Diet, data$Positive) #p = 0.16, chi-squared = 1.94 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hypotheses   Fit multivariate regression model [CODESPLIT] moltenformula <- as.formula(\"value ~ v - 1 + v:scipopgoertz + v:age + v:sex + v:lingreg_D + v:lingreg_I + v:urbanity_log + v:edu_uni + v:edu_com + v:sciprox_score + v:sciliteracy + v:rel + v:pol + v:inter_science + v:trust_science + v:trust_scientists + v:scimedia_att + v:scimedia_sat\") m_inx <- svyglm(moltenformula, design = svydsgn_melt_inx) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "estimate bifactor model (maximum likelihood) [CODESPLIT] fit <- cfa(model, dat,estimator = \"MLM\", std.lv = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "draw random factor scores from a multivariate normal distribution [CODESPLIT] lat_scores <- mvrnorm(n = n, mu = mu, Sigma = sigma) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function to create data summaries in figures [CODESPLIT] data_summary <- function(x) { m <- mean(x) ymin <- m-ci(x) ymax <- m+ci(x) return(c(y=m,ymin=ymin,ymax=ymax)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "to get numbers for confidence intervals click on item in global environment and expand plot 95% credible intervals [CODESPLIT] x<-plot(me, plot = FALSE)[[1]] + scale_color_grey() + scale_fill_grey() p<-x+ylim(0,1)+ theme(panel.grid.major = element_line(colour=\"gray\"), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_blank(),panel.grid.major.x = element_blank())+xlab(\"\")+ylab(\"\") p ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " 1  Never, 2  I tried it but don't do it systematically, 3  I do it when it feels convenient, 4  I do it for most research projects/studies, 5  I do it for every research project/study.  Make dataframe 'Wish' with intentions to engage in open science practices  [CODESPLIT] wish = data[, 17:25] wish_complete = wish[,-9] # remove missing data wish = apply(wish, 2, as.numeric) wish = as.data.frame(unlist(wish)) colnames(wish) = c(\"preregister\", \"sharedatas\", \"sharecode\", \"openaccess\", \"preprint\", \"openpeer\", \"opensw\", \"replicate\", \"other\") wish = t(wish) rowMeans(wish, na.rm = T) # mean apply(wish, 1, sd, na.rm = T) # standard deviation ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "generates 95% confidence intervals for each beta coefficient [CODESPLIT] m.stakes_decision.CI <- round(confint(m.stakes_decision, parm = \"beta_\"), 3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "2.3 Loss tangents   We draw a diagram of loss tangents with logarithmic scale on the xaxis [CODESPLIT] plot(graphData$Freq[graphData$site == \"wreck\"], graphData$tan[graphData$site == \"wreck\"], log = \"x\", axes = FALSE, xlab = \"\", ylab = \"\", type = \"b\", pch = PCH, col = col_wreck, xlim = c(0.01, 100), ylim = c(0, 15)) par(new = TRUE, ps = PS, mar = MAR) plot(graphData$Freq[graphData$site == \"beach\"], graphData$tan[graphData$site == \"beach\"], type = \"b\", log = \"x\", xlab = expression(\"Frequency [rad s\"^-1 *\"]\"), ylab = expression(paste(\"tan \", delta)), pch = PCH, col = col_beach, axes = FALSE, xlim = c(0.01, 100), ylim = c(0, 15)) axis(1, at = c(0, 0.1, 1, 10, 100, 1000), labels = c(0, 0.1, 1, 10, 100, 100)) axis(2, at = c(-1 , 0, 3, 6, 9, 12, 15)) text(x = 0.01, y = 15, label = expression(bold(\"c\"))) abline(1, 0) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "estimate mean average difference between observed and corrected values   check for sign reversals [CODESPLIT] dat_analysis[sign(MA_ES_self_corr) == sign(MA_ES_self_obs), .N] # 46 (no sign rev) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Step 3: Fit varyingintercept ESEM model [CODESPLIT] esem_anti_fit_alt1F <- cfa(esem_anti_model_alt1F, esem_anti_data_alt, group = \"country_iso\", estimator = \"MLM\", group.equal = \"loadings\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "performing the chisquared comparisons across all the possible pairs [CODESPLIT] pairwise.table(chi_squared_paired_comparison_1, rownames(hypotheses_per_research_area), p.adjust.method=\"none\") pairwise.table(chi_squared_paired_comparison_2, rownames(hypotheses_per_research_area), p.adjust.method=\"none\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "chi squared test of criteria met by association type [CODESPLIT] f_statistic <- chisq.test(data_hypotheses_long$association_type, data_hypotheses_long$reporting_criteria_met) f_statistic ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hastags   wordcloud of most common hashtags [CODESPLIT] frequenthashes <- hashtable[hashtable>floor(quantile(hashtable,.999))] hashvec <- unlist(lapply(frequenthashes,function(x) rep(names(x),x))) stretched <- log10(frequenthashes-min(frequenthashes)+1)/max(log10(frequenthashes-min(frequenthashes)+1)) cols <- rgb(1- stretched,0,0) png(\"hashcloud.png\",width = 12,heigh=12,units = \"cm\",res=300) wordcloud(names(frequenthashes),frequenthashes,random.order = F,colors=cols) dev.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit the model using unweighted least square (ULS) [CODESPLIT] m1.fit <- sem( m1, sample.cov = S, sample.nobs = 200, estimator = \"ULS\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Remove items that do not occur in the country [CODESPLIT] items <- items[items$item %in% colnames(x_country), ] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "multilevel model (repeated measures) correlation using package rmcorr: [CODESPLIT] td_corr <- rmcorr(subject, trust, distrust, td) td_corr ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Robust growth model MLR estimator [CODESPLIT] rctBoot <- growth(rctModel, data=rctWide, se = \"bootstrap\") summary(rctBoot) rctMLR <- growth(rctModel, data=rctWide, estimator = \"MLR\") summary(rctMLR) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a scatterplot of sleep quality and insomnia scores (FIRST). Add appropriate labels [CODESPLIT] plot(Sleep_and_COVID19$PSQI_global, Sleep_and_COVID19$FIRST, main = \"Main title\", xlab = \"X axis title\", ylab = \"Y axis title\", frame = FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "4) Ttest    Examine whether there is a significant difference in sleep quality in those who may have symptoms or been diagnosed with COVID [CODESPLIT] t.test(PSQI_global ~ SymptomsOrPositive, data = Sleep_and_COVID19) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Examine whether there is a significant difference in mean sleep difficulty scores in those who undertake shiftwork [CODESPLIT] t.test(PSQI_global ~ ShiftWork, data = Sleep_and_COVID19) ?t.test ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  6) Regression Analysis    Using multiple regression analysis, examine the influence of the age, hours worked, physical health, coffee drank  alertness and insomnia proneness on sleep quality  [CODESPLIT] model1 <- lm(PSQI_global ~ Age + WorkHour + Health + CoffeeToday + StanfordSleepinessScale + FIRST,  data = Sleep_and_COVID19) summary(model1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Draw samples setting heart_rate_traj above as mean with appropriate constant variance (10 beats/min) [CODESPLIT] heart_rate <- rnorm(n = length(heart_rate_traj), mean = heart_rate_traj, sd = 10) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Computing pvalues based on the Fdistribution [CODESPLIT] p_durable <- pf(q=Femp_durable, df1=Df_aest_durable, df2=Df_residual, lower.tail = F) p_nond <- pf(q=Femp_nond, df1=Df_aest_nond, df2=Df_residual, lower.tail = F) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Logistic Regression Model  Creating null model and a logistic regression model based on all variables [CODESPLIT] lr.null <- glm(Treatment ~ 1, data = dat2, family = binomial(link=\"logit\")) lrm <- glm(Treatment ~ Sq + Ssk + Sku + Sp + Sv + Sz + Sa,data = dat2, family = binomial(link=\"logit\")) summary(lrm) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Filter and select variables of interest [CODESPLIT] d <- d_2011 %>% subset(sns == 1) %>% # indicated to use social media subset(!is.na(beh_disclose_soc)) %>% select(id, country, year, gender, age, cons_record, cons_zweck, cons_target, contains(\"soc\"), v143:v156, # beh_setting, int_home, int_work, sns, nth_arg) %>% mutate(cons_mean = (cons_record + cons_zweck + cons_target)/3, cons_mean_bin = ifelse(cons_mean > median(cons_mean, na.rm = T), 1, ifelse(is.na(cons_mean), NA, 0))) %>% mutate(age_group = ifelse(age <= 32, \"younger than 32 years\", \"older than 32 years\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Confirmatory factor analyses  CFA for privacy concerns (tauequivalent) [CODESPLIT] cfa.model <- \" priv_con =~ 1*cons_record + 1*cons_zweck + 1*cons_target \" fit <- cfa(cfa.model, d_2011) summary(fit, std = T, fit = T) reliability(fit) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Estimate models across all specifications  Customized model functions to include random country effect [CODESPLIT] linear <- function(formula, data) { pb$tick() # set tick for progress bar lmer(paste(formula, \"+ (1|country)\"), data = data) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Decomposing variance  Estimate multilevel model (without predictors) [CODESPLIT] m_lin <- lmer(stdcoef ~ 1 + (1|x) + (1|controls) + (1|subsets)  + (1|x:controls) + (1|x:subsets) + (1|subsets:controls), data = results) summary(m_lin) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run polynomial regresion as a OLS linear model [CODESPLIT] f <- paste0(DV, \" ~ \", paste(\"1\", IV1, IV2, IV12, IV_IA, IV22, sep=\" + \"), controlvariables) lm.full <- lm(f, data=df, na.action=na.exclude) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Matrix of phenotypic change trajectories X and Z,  interlineage correlation matrix C, and intertrait crossproduct matrix A,  as well as the number of lineages n ( 13)  and dimensionality of the vectors p ( 76)  [CODESPLIT] X <- f.mean.diff(data.sc, fishID) Z <- X / sqrt(diag(tcrossprod(X))) C <- tcrossprod(Z) A <- crossprod(Z) n <- nrow(Z) p <- ncol(Z) - 4 # -4 accounting for the loss of d.f. from Procrustes alignment ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Histograms comparing dissimilarity between the original and  bootstrap replicates between \"Moo\" and \"Pac\" (largest and smallest dispersion)  This shows that the heteroscedasticity is real (not due to visual distortion)  [CODESPLIT] hist(acos(Z[10, ] %*% Zb[10, , ]), breaks = seq(0, pi / 2, 0.1), col = \"#FF000080\", main = \"Pac (red) versus Moo (blue)\", xlab = \"Angle between original and bootstrap vectors (radian)\") hist(acos(Z[7, ] %*% Zb[7, , ]), add = TRUE, breaks = seq(0, pi / 2, 0.1), col = \"#0000FF80\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Use kNNdistplot(d.cor.umap, k  3) to determine the value for eps  [CODESPLIT] d.cor.dbscan <- dbscan(d.cor.umap, minPts = 3, eps = 1.8) data.table.large.cluster <- data.table[,d.cor.dbscan$cluster == 1] large.cluster.cor.metric <- as.dist(acos(cor(data.table.large.cluster))) large.umap <- umap( large.cluster.cor.metric, min_dist = 1, n_neighbors = 3 ) large.dbscan <- dbscan(large.umap, minPts = 3, eps = 1) large.cols <- rainbow(length(unique(large.dbscan$cluster)), start = 2/6) write.csv( d.cor.umap, 'd_acos_cor_fixed.csv', row.names = F ) write.csv( d.cor.dbscan$cluster, 'dbscan_clustering_fixed.csv' ) write.csv( large.umap, 'd_acos_cor_fixed_large_cluster.csv', row.names = F ) write.csv( large.dbscan$cluster, 'dbscan_clustering_of_large_cluster_fixed.csv' ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Chisquare test: Difference in preference based on outcome, by condition [CODESPLIT] chisq.test(matrix(c(19,5,5,17), ncol = 2, byrow = T)) oddsratio.wald.outcome <- oddsratio.wald(matrix(c(19,5,5,17), ncol = 2, byrow = T))$measure[2] # Note: I reorganized the matrix so that r is of the same sign as that of Experiment 2 esc_2x2( grp1yes = 19, grp1no = 5, grp2yes = 5, grp2no = 17, es.type = \"or\" ) # this confirms Wald's odds ratio ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Two sample ttest: Difference in proportionate looking based on outcome, by condition [CODESPLIT] twosamplet.outcome <- t.test(pPositiveO ~ Condition, data = bh2zoom, alternative = \"two.sided\") twosampled.outcome <- cohen.d(pPositiveO ~ Condition, data = bh2zoom)$estimate ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Assess how well model captured PWLs ' ' This function calculates several indicators that help to assess how well the model captured captWKLs beyond simple layer structure. ' It get's called by the overarching evaluation script `evaluate_main.R` ' ' So, for each WKL the function computes ' ' * correlation factors for likelihood/distribution/sensitvity versus proportion of unstable grid points ' * the maximum proportion of unstable grid points during the lifetime of the WKL ' * the temporal lag between first forecaster concern (> avy problem) and first time that grid points are unstable in that layer '      different timing instances possible: first time (i) at least one grid point unstable, (ii) the majority of grid points is unstable, (iii) more than 50 % of max proportion unstable ' * the temporal lag between latest forecaster concern (> last day of avy problem) and last time that grid points are unstable in that layer '      need to find rules for becoming dormant and waking up again versus becoming inactive ' ' @param VData object;; don't include both primary/secondary and tertiary gtype_ranks! ' ' @export  [CODESPLIT] assessQualityOfcaptWKL <- function(VData, band, stabilityindex = \"p_unstable\") { gtype_rank <- unique(VData$vframe$gtype_rank) if (any(c(\"primary\", \"secondary\") %in% gtype_rank) & \"tertiary\" %in% gtype_rank) stop(\"You can only provide either primary and/or secondary, or tertiary gtype_ranks in your VData object to this function!\") possible_ranks <- c(\"tertiary\", \"secondary\", \"primary\") gtype_rank <- possible_ranks[possible_ranks %in% gtype_rank][1] nrow_max <- length(VData$wkl$wkl_uuid) OUTnames <- c( \"wkl_uuid\", \"wkl_iscrust\", \"wkldate\", \"wkltag\", \"nPDays\", \"nPDays_anyunstable\", \"nPDays_median\", \"nPDays_halfofmax\", \"nPDays_20\", \"rho_llhd\", \"rho_dist\", \"rho_sens\", \"p_llhd\", \"p_dist\", \"p_sens\", \"pu_max\", \"offset_maxes\", \"offset_mean\", \"pcapt_max\", \"lagA_anyunstable\", \"lagA_median\", \"lagA_halfofmax\", \"lagA_20\", \"lagZ_anyunstable\", \"lagZ_median\", \"lagZ_halfofmax\", \"lagZ_20\", \"band\", \"stabilityindex\", \"gtype_rank\" ) OUT <- matrix(nrow = max(1, nrow_max), ncol = length(OUTnames), dimnames = list(seq(max(1, nrow_max)), OUTnames)) for (i in seq_along(VData$wkl$wkl_uuid)) { wuid <- VData$wkl$wkl_uuid[i] wkldate <- as.character(as.Date(VData$wkl$datetag[i])) wkltag <- paste(format(as.Date(VData$wkl$datetag[i]), \"%b %d\"), substr(gtype_rank, start = 1, stop = 4)) wkl_iscrust <- as.logical(VData$wkl$iscrust[i]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Correlation of life satisfaction and depression across all measurement occasions [CODESPLIT] d.all %>% select(contains(\"fsat\"), contains(\"depr\"), id) %>% gather(key, value, -id) %>% separate(key, c(\"time\", \"variable\")) %>% spread(variable, value) %>% select(depr, fsat) %>% zero_order_corr(print = T, digits = 3) %>% select(`1`) %>% slice(2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check if latent covariances are equal across groups [CODESPLIT] est_svc <- cfa(mod_s,quop_use, estimator = \"MLR\", missing = \"FIML\", group = \"year\", group.equal = c(\"loadings\",\"intercepts\",\"residuals\", \"lv.variances\",\"lv.covariances\"), cluster = \"class\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get factor scores from model est_svcm_sl [CODESPLIT] fs <- lavPredict(est_svcm_sl, method = \"bartlett\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  Plot for distribution of emotions   Arguments: original ddbb, type of emotion, plot title  [CODESPLIT] create.emotion.plot <- function(case_ddbb, emotion=\"emotion\", plot_title=NULL){ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  Table and plot for effect sizes   Arguments: original ddbb, plot title  [CODESPLIT] create.effectsize.tableplot <- function(case_ddbb, plot_title = NULL){ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert average time into a ordinal variable [CODESPLIT] Tbl2$DurSecAvg1Cat <- cut(Tbl2$DurSecAvg1, breaks = quantile(Tbl2$DurSecAvg1, c(0, 1/3, 2/3, 1)), include.lowest = T) aggregate(Tbl2$DurSecAvg1, list(Tbl2$DurSecAvg1Cat), summary) levels(Tbl2$DurSecAvg1Cat) <- c(\"Less\", \"Avg\", \"More\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plotting Figure 3 [CODESPLIT] png(filename = \"Fig03_Demographics.png\", width = 19, height = 10, units = \"cm\", res = 500, pointsize = 7) par(mfrow=c(2,3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Predict node for checking differences in distributions with chisquared test [CODESPLIT] Tbl2$NodeRate <- predict(Engage, newdata = Tbl2, type = \"node\") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$FeedbackEngage)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) Tbl2$NodeRate <- predict(FeedTree, newdata = Tbl2, type = \"node\") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$Useful_FeedbackMiddle)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) Tbl2$NodeRate <- predict(ExTree, newdata = Tbl2, type = \"node\") table(Tbl2$NodeRate) (temp <- chisq.test(Tbl2$NodeRate, Tbl2$Useful_FeedbackMiddle)) temp$observed round(100*prop.table(temp$observed, 1), 1) mosaic(temp$observed, shade = T) head(Tbl2$RankTaskPerform) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert `from_name` to factor with three levels. (Facilitates data visualization). [CODESPLIT] Campaign_Messages$from_name <- factor(Campaign_Messages$from_name, levels = c (\"Stephen Harper\", \"Justin Trudeau\", \"Tom Mulcair\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "change the title text to size 20 and bold [CODESPLIT] axis.title = element_text(size = 20, face = \"bold\"), axis.title = element_text(size = 20, face = \"bold\"), ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " change the axis label text to size 14, bold, and black color  [CODESPLIT] axis.text.x = element_text(size = 14, face = \"bold\", color = \"black\"), axis.text.y = element_text(size = 14, face = \"bold\", color = \"black\")) + axis.text.x = element_text(size = 14, face = \"bold\", color = \"black\"), axis.text.y = element_text(size = 14, face = \"bold\", color = \"black\")) + scale_y_continuous(limits = c(0, 17), expand = c(0,0)) + ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Identify all neighbours within 2km [CODESPLIT] sp_nb <- spdep::dnearneigh(coords, d1 = 0, d2 = km, row.names = row.names(sp), longlat = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "take the inverse of the distances [CODESPLIT] W <- lapply(dsts, function(x) 1 / x) W <- lapply(dsts, function(x) 1 / x) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "recode base price and overage as numeric (without $ sign) [CODESPLIT] x$b1=as.numeric(sapply(strsplit(as.character(x$b1),\"\\\\$\"),function(y) y[2])) x$b2=as.numeric(sapply(strsplit(as.character(x$b2),\"\\\\$\"),function(y) y[2])) x$b3=as.numeric(sapply(strsplit(as.character(x$b3),\"\\\\$\"),function(y) y[2])) x$c1=as.numeric(sapply(strsplit(as.character(x$c1),\"\\\\$\"),function(y) y[2])) x$c2=as.numeric(sapply(strsplit(as.character(x$c2),\"\\\\$\"),function(y) y[2])) x$c3=as.numeric(sapply(strsplit(as.character(x$c3),\"\\\\$\"),function(y) y[2])) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " create dataset with possible k values, their prior and likelihood  probabilities  likelihood is drawn from 1000 random samples from beta distibution  [CODESPLIT] df <- tibble(k = hypo_k) %>% mutate(model_vote = 0.0727^k) %>% mutate(occur = map_int(.$model_vote, ~ length(which(polls$value >= .x))), prior = prob_k) %>% mutate(like = map(.$occur,~ rbeta(trials, .x, polls_len - .x))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "visualise normalised posterior on a graph [CODESPLIT] p <- df %>% ggplot(aes(k, norm_posterior)) + geom_point() + geom_line() p + labs( title = \"Probability of k exponent in the interval <0.89;; 0.99>\", x = \"Value o k exponent\", y = \"Normalised probability\" ) + theme_minimal() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "3) CALCULATION OF ESTIMATED MARGINAL MEANS   Tests for differences in response to jargon for plot [CODESPLIT] at_jarg <- list(Jargon = c(\"More\", \"Less\"), UnderN = \"Easy\", RecognN = \"Fairly\", TrvlAdv_Atten = \"Cons\", cut = \"3|4\") grid_jarg <- ref_grid(clmm_usejar, mode = \"exc.prob\", at = at_jarg) (emm_jarg <- emmeans(grid_jarg, specs = pairwise ~ Jargon, by = \"BackgrAvTraining\")) plot_jarg <- summary(emm_jarg$emmeans) cont_jarg <- summary(emm_jarg$contrasts) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Gradient of FDR with respect to alpha (and its implied power) across ncp for a onesided ztest  Note that the gradient is increasing towards $\\infty$ from both side at ncp  0 and alpha 0, but is equal to 0 at that point. The figure does not allow to depict it properly and it seems like as if it peaked at ncp around .5 and was decreasing towards 0.  Green corresponds to $P(H_0)  .2$, blue to $P(H_0)  .5$, and red to $P(H_0)  .8$.  [CODESPLIT] alpha <- seq( 0, 1, length.out = 500) ncp   <- seq(-5, 5, length.out = 500) FDR.2od  <- fill_gradFDR.alpha(.2, alpha, ncp, FALSE) FDR.5od  <- fill_gradFDR.alpha(.5, alpha, ncp, FALSE) FDR.8od  <- fill_gradFDR.alpha(.8, alpha, ncp, FALSE) plot_ly(x = ~alpha, y = ~ncp, z = ~FDR.5od) %>% add_surface(contours = list( y = list( show = TRUE, project = list(y = TRUE), usecolormap = FALSE, color = \"blue\" ), x = list( show = TRUE, project = list(x = TRUE), usecolormap = FALSE, color = \"blue\" ) ), opacity = .4) %>% layout(scene = list( xaxis = list( tickvals = c(0, .05, .1, .2, .5, 1) ), yaxis = list( tickvals = pretty(ncp) ), zaxis = list( tickvals = seq(0, 1, .2), title    = list(text = \"FDR'\") ) )) %>% add_surface(x = ~alpha, y = ~ncp, z = ~FDR.8od, opacity = .4, contours = list( y = list( show = TRUE, project = list(y = TRUE), color = \"red\" ), x = list( show = TRUE, project = list(x = TRUE), color = \"red\" ) )) %>% add_surface(x = ~alpha, y = ~ncp, z = ~FDR.2od, opacity = .4, contours = list( y = list( show = TRUE, project = list(y = TRUE), color = \"green\" ), x = list( show = TRUE, project = list(x = TRUE), color = \"green\" ) )) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "List that will include table (each element should be vector of 3): [CODESPLIT] Table <- list( c(\"Network Characteristics\",title,\"\"), c(\"Comparing Global Characteristics\",nameOrig, nameRepl) ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " The row with \"def_cond\" is moved to a new column and duplicated (this is  done with reference to \"cond_to_group\")  [CODESPLIT] df_tmp_2 <- full_join( df_tmp_1 %>% filter({{cond_to_compare}} != def_cond), df_tmp_1 %>% filter({{cond_to_compare}} == def_cond), by = col ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Repeating the analyses with closeness as covariate [CODESPLIT] cor.test(data_3A$closeness_1,data_3A$closeness_2) cor.test(data_3B$closeness_1,data_3B$closeness_2) cor.test(data_3C$closeness_1,data_3C$closeness_2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check if the number of processed partners is smaller than the number of listed partners as saved in the counter variable [CODESPLIT] nrow(Relationship_details) < My_EHC_survey$count_SP", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run anovas to check interaction between specificity and condition  LA [CODESPLIT] anova_N4_la_sp = summary(aov(Voltage ~ Condition *Specificity , data = mydata_n4_la)) anova_N4_la_sp ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Visualize grid cluststers [CODESPLIT] mgp.u<-c(2.3,0.6,0.5) if(tifit==T){ tiff(tifname,width=14,height=12.5,units=\"cm\",res=600,compression=\"lzw\") } par(oma=c(0,4,4,4)) layout(matrix(1:9,ncol=3,byrow=T),widths=c(1.28,1,1),heights=c(1,1,1.28)) doFloorC(n=1000,eta=1,char=char,mar1=c(gap,3.5,gap,gap),mgp1=mgp.u,xax=F,yax=T,xl=\"\",col2=col2,seq.s=seq.s,seq.c=seq.c,contour=contour) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create the dataset for visualizations Summarize the means (and sd for clusters) [CODESPLIT] means.cl<-tapply(res$clust,paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) sd.cl<-tapply(res$clust,paste(res$n,res$h,res$nu,res$eta),sd,na.rm=T) means.dim<-tapply(res$explained3D,paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) means.avg<-tapply(log(res$avgdist),paste(res$n,res$h,res$nu,res$eta),mean,na.rm=T) vars<-strsplit(names(means.cl),\" \") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Graph of the distribution of pvalues for each effect [CODESPLIT] pvalue.plot <- ggplot(melt_pvalue,aes(x = value)) + facet_wrap(~variable, ncol = 5) + geom_histogram() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Graph of the distribution of standard estimates for each effect [CODESPLIT] est.std.plot <- ggplot(melt_est.std,aes(x = value)) + facet_wrap(~variable, ncol = 5) + geom_histogram() list.result <- list(\"dist_pvalue.plot\" = pvalue.plot, \"dist_est.std.plot\" = est.std.plot) return(list.result) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "returns plot of pvalue disttribution, plot of standard estimates distribution  [CODESPLIT] list.result <- list(\"dist_pvalue.plot\" = pvalue.plot, \"dist_est.std.plot\" = est.std.plot) return(list.result) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  Model estimates plot   Arguments: standard estimates distribution, pvalues distributions, use sign level TREU/FALSE, level of signicance, plot title  [CODESPLIT] create.model_estimates.plot <- function (est.std_dist, pvalue_dist, significant = FALSE, sig_level=.05, title = NULL){ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Median of each standard estimates and TRUE/FALSE values if mean(pvalues) < sig_level [CODESPLIT] est.std_All <- data.frame(cbind(colMeans(est.std_dist), colMeans(pvalue_dist) <= sig_level)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Change the participantID to a shorter & uniform name \"Subject\"  [CODESPLIT] names(d)[names(d) == \"Participant.Public.ID\"] <- \"Subject\" ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " For trimmed set, we remove any RT that is too long (>2500ms)  [CODESPLIT] d_trimmed = d %>% filter (acc == 2, Reaction.Time > 300, Reaction.Time < 2500) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plotting   create a dataset that contains means of both oddnumbered & evennumbered items' RTs [CODESPLIT] raw_untrimmed<-inner_join(x= mean.o, y=mean.e, by=\"Subject\") names(raw_untrimmed) <- c(\"Subject\", \"Odd\", \"Even\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "AVERAGE BY PARTICIPANT [CODESPLIT] mean.e = de %>% group_by(Subject) %>% summarise(mean(rt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(rt.diff)) mean.e = de %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.e = de %>% group_by(Subject) %>% summarise(mean(zrt.diff)) mean.o = do %>% group_by(Subject) %>% summarise(mean(zrt.diff))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "combine them into one data frame [CODESPLIT] person <- inner_join(person_rand_e, person_rand_o, by = \"Subject\", suffix = c(\"_even\", \"_odd\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Wilcoxon ranksum test on widths of unbroken artefacts: [CODESPLIT] results <- wilcox.test(An2entire$Width,Gmeentire$Width) Z       <- qnorm(results$p.value) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Wilcoxon ranksum test on lengths of unbroken artefacts: [CODESPLIT] results <- wilcox.test(An2entire$Length,Gmeentire$Length) Z       <- qnorm(results$p.value) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "DATA DOWNSAMPLING  add collumns with delta distance and delta time for downsampling [CODESPLIT] trajectory.df <- trajectory.df %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = \"min\"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Caculate daily cumulative distances and time lags Calculate speed Join with relocation info Add consecutive number per individual to indicate tracking day [CODESPLIT] daily <- trajectory.df %>% group_by(id, date) %>% filter(!dist == \"NA\") %>% dplyr::summarize(daily_dist = sum(dist), sex = first(sex), time_lag_min = sum(time_lag_min)) %>% mutate(id_day = paste(id, date), daily_speed = (daily_dist/as.numeric(time_lag_min) * 60)) %>% left_join(relocs.perday) %>% group_by(id) %>% mutate(trans_day = row_number()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Split up the trajectories into 50m and 200m translocations [CODESPLIT] trajectory50m <- trajectory.df %>% filter(trans_group == \"50m\") trajectory200m <- trajectory.df %>% filter(trans_group == \"200m\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot daily movement by daily temp & sex [CODESPLIT] ggplot(data = daily, aes(x = temp, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot daily movement by daytime rainfall & sex [CODESPLIT] ggplot(data = daily, aes(x = rain_daytime, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot daily movement by cumulative rainfall & sex [CODESPLIT] ggplot(data = daily, aes(x = rain_cumul, y = daily_dist, group = sex, color = sex)) + geom_point() + stat_smooth(method=lm) + theme_bw() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " x is a matrix containing the data  method : correlation method. \"pearson\"\" or \"spearman\"\" is supported  removeTriangle : remove upper or lower triangle  results :  if \"html\" or \"latex\"  the results will be displayed in html or latex format  [CODESPLIT] corstars <-function(x, method=c(\"pearson\", \"spearman\"), removeTriangle=c(\"upper\", \"lower\"), result=c(\"none\", \"html\", \"latex\")){ ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Diagonal matrix Tau used to scale the z_w [CODESPLIT] Tau_w <- matrix(c(DA.mp$`tau_w[1]`,0,0, 0,DA.mp$`tau_w[2]`,0, 0,0,DA.mp$`tau_w[3]` ),nrow=3,ncol=3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Run fixed effect model [CODESPLIT] life <- rma(yi = corrs$cor, vi = corrs$vi, measure = \"COR\", method = \"FE\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Set max Y for graph [CODESPLIT] y_max<-max(d_dist)+1", "target": 1, "target_options": ["no_match", "match"]}
{"input": " This does the summary. For each group's data frame, return a vector with  N, mean, median, and sd  [CODESPLIT] datac <- ddply(data, groupvars, .drop=.drop, .fun = function(xx, col) { c(N    = length2(xx[[col]], na.rm=na.rm), mean = mean   (xx[[col]], na.rm=na.rm), median = median (xx[[col]], na.rm=na.rm), sd   = sd     (xx[[col]], na.rm=na.rm) ) }, measurevar ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Rename the \"mean\" and \"median\" columns  [CODESPLIT] datac <- rename(datac, c(\"mean\" = paste(measurevar, \"_mean\", sep = \"\"))) datac <- rename(datac, c(\"median\" = paste(measurevar, \"_median\", sep = \"\"))) datac$se <- datac$sd / sqrt(datac$N)  # Calculate standard error of the mean ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " A function that reads in one line in the csv file, add tags and texts around it, and write the output line into an output file  [CODESPLIT] add_tags <- function(l,o){ image_no <- strsplit(as.character(l[1]),\"\\\\.\")[[1]][1] image_URL <- paste(\"<img src='\", as.character(l[2]),\"'>\",sep=\"\") write(paste(\"[[Block:\", image_no, \"]] \\n [[Question:DB]] \\n [[ID:\", image_no, \"-image]] \\n\", image_URL, \" \\n [[Question:TE:SingleLine]] \\n [[ID:\", image_no, \"-TE]] \\n\", TE, \" \\n [[Question:MC:SingleAnswer:Vertical]] \\n [[ID:\", image_no, \"-MC]] \\n\", MC, \" \\n [[Choices]] \\n\", choices_formatted, \"\\n [[Question:Matrix]] \\n [[ID:\", image_no, \"-ratings]] \\n \",rating_prompt,\" \\n [[AdvancedChoices]] \\n \", rating_statements_formatted, \"\\n [[AdvancedAnswers]] \\n [[Answer]] \\n\",\"1 - Strongly Disagree\", \"\\n [[Answer]] \\n 2 \\n [[Answer]] \\n 3 \\n [[Answer]] \\n 4 \\n [[Answer]] \\n 5 \\n [[Answer]] \\n 6 \\n [[Answer]] \\n\", \"7 - Strongly Agree\",\" \\n [[PageBreak]] \\n\", sep = \"\"), o, append = TRUE) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate a plot for posterior predictive check (evaluate whether the ' posterior predictive data look more or less similar to the observed data) ' ' @param df dataframe with the data ' @param mod Bayesian model ' ' @return plot with posterior predictive check [CODESPLIT] get_pp_check <- function(df, mod) { map(list(df), ~brms::pp_check(mod, ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Simulate a single between group study given sample size n, true (fixed) effect size delta, and heterogeneity (random effect) tau.  Generate data for the experimental group (y_e) and for the control group (y_c).  [CODESPLIT] y_e = rnorm(n, 0, 1) + delta + rnorm(1, 0, tau) y_c = rnorm(n, 0, 1) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " calculate pooled variance S, standardized mean difference d, the variance of d, the pvalue, and N.  [CODESPLIT] S = sqrt(((n - 1) * v_e + (n - 1) * v_c) / df) d = (m_e - m_c) / S var.d = (n + n)/(n * n) + (d^2 / (2 * (n + n))) se.d = sqrt(var.d) N = n + n", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Box and Whisker Plots to Compare Models [CODESPLIT] scales <- list(x=list(relation=\"free\"), y=list(relation=\"free\")) bwplot(results, scales=scales) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "stimuli  specify correlations for rnorm_multi (one of several methods) [CODESPLIT] stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c(\"stim_i\", \"stim_version_slope\") ) %>% mutate( stim_id = 1:stim_n ) stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c(\"stim_i\", \"stim_version_slope\") ) %>% mutate( stim_id = 1:stim_n ) stim_cors = stim_i_cor stim = rnorm_multi( n = stim_n, vars = 2, r = stim_cors, mu = 0, # means of random intercepts and slopes are always 0 sd = c(stim_sd, stim_version_sd), varnames = c(\"stim_i\", \"stim_version_slope\") ) %>% mutate( stim_id = 1:stim_n ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate trialspecific effects by adding overall effects and slopes [CODESPLIT] version_eff = stim_version_eff + stim_version_slope + sub_version_slope, version_eff = stim_version_eff + stim_version_slope + sub_version_slope, version_eff = stim_version_eff + stim_version_slope + sub_version_slope, ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Take the list of csv files from above, and read them all into R (purrr::map function)  This will download 100 csv files into a list  The reduce function will then bind them all together into a dataframe  [CODESPLIT] large_df <- csv_files %>% purrr::map(function(x) { read.csv(x) }) %>% purrr::reduce(cbind)", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Re structure this dataframe to turn it to long format (pivot_longer)  Keep the X columns in, these are the rowvalues we can use to denote a sample  The column names contain the participant (P) and the session number (S)  It will display with each 1st sample of each session for a given participant, so need to reorder to get sorted by participant, session, then second sample  The names_pattern uses regular expresisons to denote that the names will come from the specified area  i.e., after the P, which could contain any values which occur before the underscore, and likewise after the S  [CODESPLIT] large_df <- large_df %>% tidyr::pivot_longer(cols = !contains(\"X\"), names_to = c(\"participant\", \"session\"), names_pattern = c(\"P(.*)_S(.*)\"), values_to = \"heart_rate\")  %>% dplyr::rename(seconds = \"X\") %>% dplyr::arrange(participant, session, seconds) %>% dplyr::mutate(participant = as.factor(participant), session = as.factor(session)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "pool \"exp_buc\" columns into one column, so that advocacy type is one variable and create new \"experienced\" column for each level of advocacy type  [CODESPLIT] data <- data %>% pivot_longer(names_to = \"advocacytype\", values_to = \"experienced\", graphic_exp_buc:disprotest_exp_buc) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create variables for residue that is 'present' vs 'absent' This is used for the tables/descriptives AND for color/filling the figures [CODESPLIT] data <- dplyr::mutate(data, oro_zero = if_else(valleculae_severity_rating == 0, \"Absent\", \"Present\")) data <- dplyr::mutate(data, hypo_zero = if_else(piriform_sinus_severity_rating == 0,\"Absent\", \"Present\")) data <- dplyr::mutate(data, epi_zero = if_else(epiglottis_severity_rating == 0, \"Absent\", \"Present\")) data <- dplyr::mutate(data, lv_zero = if_else(laryngeal_vestibule_severity_rating == 0, \"Absent\", \"Present\")) data <- dplyr::mutate(data, vf_zero = if_else(vocal_folds_severity_rating == 0, \"Absent\", \"Present\")) data <- dplyr::mutate(data, sg_zero = if_else(subglottis_severity_rating == 0, \"Absent\", \"Present\")) data <- dplyr::mutate(data, pas_zero = if_else(pas_max < 2, \"Absent\", \"Present\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Diagnostic tests to determine appropriate number of factors in EFA  parallel analysis (includes scree plot) and Very Simple Structure test  optimal number of factors is 6 [CODESPLIT] fa.parallel(mainData[, c(2:56)], fm = 'ml', fa = 'fa') vss(mainData[, c(2:56)], n = 8, rotate = 'oblimin', fm = 'mle') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "replace missings values (9 and 99) with NA [CODESPLIT] pilot<- na_if(pilot,99) pilot<- na_if(pilot,-9) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "subset of participants with less then 4 missings [CODESPLIT] pilot<- subset(pilot, pilot$missings_a < 4) pilot_c <- pilot[,c(\"SJT_kb_00007\",\"SJT_kb_00028\",\"SJT_kb_00058\", \"SJT_kb_00054\",\"SJT_kb_00072\",\"SJT_kb_00026\", \"SJT_kb_00060\",\"SJT_kb_00053\",\"SJT_kb_00039\", \"SJT_kb_00027\",\"SJT_kb_00015\",\"SJT_kb_00038\", \"SJT_kb_00035\",\"SJT_kb_00055\",\"SJT_kb_00010\")] pilot$missings_c <- apply(pilot_w,1,function(x) sum(is.na(x))) pilot<- subset(pilot, pilot$missings_c < 4) pilot_es <- pilot[,c(\"SJT_kb_00103\",\"SJT_kb_00197\",\"SJT_kb_00136\", \"SJT_kb_00134\",\"SJT_kb_00113\",\"SJT_kb_00128\", \"SJT_kb_00117\",\"SJT_kb_00127\",\"SJT_kb_00104\", \"SJT_kb_00108\",\"SJT_kb_00094\",\"SJT_kb_00143\", \"SJT_kb_00116\",\"SJT_kb_00202\",\"SJT_kb_00175\")] pilot$missings_es <- apply(pilot_es,1,function(x) sum(is.na(x))) pilot<- subset(pilot, pilot$missings_es < 4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Coding past experiences into binary variables [CODESPLIT] data <- data %>% mutate_at(vars(\"graphic_exp\":\"attentioncheck_exp\"), list(buc = bucketing_experienced)) %>% glimpse ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Coding behaviors into binary variables 0  unselected, 1  selected  [CODESPLIT] data <- data %>% mutate_at(vars(\"behavior_donation\":\"behavior_none\"), list(bin = binary.behavior)) %>% glimpse ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Creating a combined race/ethnicity variable [CODESPLIT] data <- data %>% mutate(race.eth = case_when(hispanic == \"Yes\" ~ \"Hispanic or Latinx\", race == \"Caucasian/White\" & hispanic == \"No\" ~ \"White, Non-Latinx\", race == \"Black or African-American\" & hispanic == \"No\" ~ \"Black, Non-Latinx\", race == \"Asian\" & hispanic == \"No\" ~ \"Asian\", race == \"Multi-Racial\" & hispanic == \"No\" ~ \"Multi-Racial\", TRUE ~ \"Other Races/Ethnicities\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "correlations between csd_mean and averaged csds for Big Five traits (not mentioned   in the text): [CODESPLIT] psych::corr.test( df2$e_csd,df2$csd_mean_e ) psych::corr.test( df2$a_csd,df2$csd_mean_a ) psych::corr.test( df2$pa_csd,df2$csd_mean_pa ) psych::corr.test( df2$na_csd,df2$csd_mean_na ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Participants could select all response options that applied to them in any given situation. Here, we create one dummy variable for every response option.  [CODESPLIT] for (i in 1:length(survey$interacting_by)) { v <- unlist(strsplit(survey$interacting_by[i], \",\")) if (is.element(\"1\", v)) { survey$TiP[i] <- 1 } if (is.element(\"2\", v)) { survey$ToP[i] <- 1 } if (is.element(\"3\", v)) { survey$TM[i] <- 1 } if (is.element(\"4\", v)) { survey$ChattingWHATSAPP[i] <- 1 } if (is.element(\"5\", v)) { survey$ChattingDATING[i] <- 1 } if (is.element(\"6\", v)) { survey$Emailing[i] <- 1 } if (is.element(\"7\", v)) { survey$Videochatting[i] <- 1 } if (is.element(\"8\", v)) { survey$Facebook[i] <- 1 } if (is.element(\"9\", v)) { survey$Instagram[i] <- 1 } if (is.element(\"10\", v)) { survey$Snapchat[i] <- 1 } if (is.element(\"11\", v)) { survey$Twitter[i] <- 1 } if (is.element(\"12\", v)) { survey$OTHER[i] <- 1 } if (is.element(\"0\", v)) { survey$NoInteraction[i] <- 1 } if (is.element(\"999\", v)) { survey$SKIP[i] <- 1 } } for (i in 1:length(survey$interacting_people)){ v <- unlist(strsplit(survey$interacting_people[i], \",\")) if (is.element(\"1\", v)) { survey$Classmates[i] <- 1 } if (is.element(\"2\", v)) { survey$Coworkers[i] <- 1 } if (is.element(\"3\", v)) { survey$Family[i] <- 1 } if (is.element(\"4\", v)) { survey$Friends[i] <- 1 } if (is.element(\"5\", v)) { survey$Roommates[i] <- 1 } if (is.element(\"6\", v)) { survey$Significant_other[i] <- 1 } if (is.element(\"7\", v)) { survey$Strangers[i] <- 1 } if (is.element(\"8\", v)) { survey$OTHER2[i] <- 1 } if (is.element(\"999\", v)) { survey$SKIP2[i] <- 1 } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Rescore so high values are what we expect to be positive predictors (e.g. female 2 male 1) [CODESPLIT] df$Nationality2 <- df$Nationality df$Nationality2 <- as.integer(df$Nationality2 %in% \"Dutch\") df$Nationality2[df$Nationality2==0] <- 2 df$Nationality2[df$Nationality2==1] <- 0 df$Nationality2[df$Nationality2==2] <- 1  # Dutch=0 (36), Rest=1 (44) count(df$Nationality2) df$RelationshipStatus2 <- df$RelationshipStatus df$RelationshipStatus2 <- recode(df$RelationshipStatus2, \"3\" = 2) df$RelationshipStatus2[df$RelationshipStatus2==2] <- 0  # Partnered=0 (30), Single=1 (50) count(df$RelationshipStatus2) df$Gender2 <- df$Gender          # df$Gender2[df$Gender2==3] <- NA  # code 1 nonbinary participant as NA for this to avoid 3 levels df$Gender2[df$Gender2==1] <- 0 df$Gender2[df$Gender2==2] <- 1   # men=0 (19), women=1 (60), 1 NA count(df$Gender2) df$WorkStatus2 <- df$WorkStatus df$WorkStatus2[df$WorkStatus2==2] <- 0 # 54no, 26yes count(df$WorkStatus2) df$MentalHealth2 <- as.numeric(df$MentalHealth) df$MentalHealth2[df$MentalHealth2==1] <- 0 df$MentalHealth2[df$MentalHealth2==2] <- 1   # healthy=0 (59), history=1 (17), 4 NA count(df$MentalHealth2) df$selfefficacy_pre2 <- df$selfefficacy_pre*-1 # reverse code ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Rating Means and Standard Deviations of Conditions   Unspecified Rating [CODESPLIT] round(mean(unspecified$rating), 2) round(sd(unspecified$rating), 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Save and export image as .eps (8.18 x 4.88 inch) [CODESPLIT] if(saveFigures) cairo_ps(file = '../R_Output/Images/AccuracyPerProject_Revision.eps', onefile = TRUE, fallback_resolution = 600, width = 8.18, height = 4.88) par(cex.main = 2.2, mar = c(5, 5.5, 2, 2) + 0.1, mgp = c(3, 1, 0), cex.lab = 2, font.lab = 1, cex.axis = 1.6, bty = 'n', las = 1) plot(x = c(.5, .8), y = c(0, 50), type = 'n', xlab = expression(paste(\"Accuracy Rate \", omega, \" (in %)\")), ylab = 'Density', main = '', axes = FALSE) axis(2, lwd.tick = 0, labels = FALSE) axis(1, at = seq(.5, .8, by = .05), labels = seq(50, 80, by = 5)) d <- lapply(split(df$omega, df$condition), density) Map(function(dens, col) polygon(dens, col = col), dens = d, col = figureColors[['cols.transparent']]) text(.64,44.5,'Laypeople\\n(Evidence)', cex = 1.3) text(.55,44.5,'Laypeople\\n(Description)', cex = 1.3) text(.621,25,'Experts\\nML2',  cex = 1.3) text(.75,25, 'Experts\\nSSRP', cex = 1.3) graphics.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " function to calculate AUC [CODESPLIT] groupAUC.description <- pnorm(description.mud / sqrt(2)) # signal distribution has sigma = 1 quantile(groupAUC.description, c(0.025, 0.5, 0.975)) groupAUC.evidence <- pnorm(evidence.mud / sqrt(2)) # signal distribution has sigma = 1 quantile(groupAUC.evidence, c(0.025, 0.5, 0.975)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get relevant information  compute mode of signal distribution [CODESPLIT] estimate_mode <- function(x) { d <- density(x) d$x[which.max(d$y)] } modeD           <- estimate_mode(mud) medianC         <- median(muc) empiricalLambda <- (0.5*modeD) + medianC", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compute CIs for each value of lambda [CODESPLIT] probD      <- quantile(mud, probs = probs) for (i in seq_len(nPoints)){ c        <- lambdas[i] - (0.5 * probD) qph[i, ] <- pnorm((0.5 * probD) - c) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " get the posterior distributions of the intercepts, transformed to rate scale  [CODESPLIT] post.description <- inv_logit_scaled(posterior_samples(m.studies, 'b_intercept'))[,1] post.evidence    <- inv_logit_scaled(posterior_samples(m.studies, 'b_intercept') + posterior_samples(m.studies, 'b_conditionDescriptionPlusEvidence'))[,1] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "add the observed average accuracy per study and per condition [CODESPLIT] fit$obs.accuracy <- c(aggregate(dat$guessed.correctly, list(condition = dat$condition), mean)[,'x'], aggregate(dat$guessed.correctly, list(study = dat$study, condition = dat$condition), mean)[,'x']) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a binary trial type variable 0  Go, 1  No Go  [CODESPLIT] sart$trial_type <- sart$code sart$trial_type[sart$code == 3] <- NA sart$trial_type <- as.factor(sart$trial_type-1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate Go RT coefficient of variation (CV) to index intraindividual variability [CODESPLIT] sartsub$go_cv <- sartsub$gosd / sartsub$gomean sart <- merge(sart, sartsub[,c(1,5,6,9)], by = \"subid\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add CSE variables to the MW probe data set [CODESPLIT] probes <- merge(probes, cse_wide[,c(\"subid\",\"cse\",\"ce\",\"csez\",\"cez\")], by = \"subid\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " read in, display, and summarize 'csv' data  [CODESPLIT] data_raw <- read.table(\"osylv_spaceuse_raw.csv\", header = T, sep =\",\") head(data_raw) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add collumns with delta distance and delta time [CODESPLIT] tracks <- tracks %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = \"min\"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "boxplot of maximum dist per id [CODESPLIT] maxdist_plot <- ggplot(daily_perid, aes(x= sex, y=max_dist_perid, fill = sex)) + theme_bw(20) + geom_boxplot(aes(fill = sex), outlier.shape = NA) + labs(y = \"Longest movement (m)\") + geom_jitter(aes(fill=sex), position=position_jitterdodge(0.2), shape = 21, size = 4, alpha=0.4) + facet_wrap(~site, labeller = labeller( site=c(can =\"Natural site\", oto = \"Enclosures\"))) + scale_x_discrete(labels= c(\"F\", \"M\")) + theme(legend.position= \"none\", axis.title.x = element_blank(), axis.ticks.x = element_blank(), axis.text.x = element_text(color = \"black\", size = 18), aspect.ratio = 3) + ylim(0, 25) # excludes one male point at >40m from Canande maxdist_plot ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "centering a variable to a range from 0.5 to 0.5 [CODESPLIT] center <- function(x) { minimum <- min(x) maximum <- max(x) return((x - (minimum + maximum)/2)/(maximum - minimum)) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "keep only data from 2003 onwards (because fid started in 2003) [CODESPLIT] survival <- subset(survival, year >2002) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " identify influenial points (e.g., cooksD > 6/n)  [CODESPLIT] influential_obs <- as.numeric(names(cooksD)[(cooksD > (6/n))])", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function serves to plot all important information what is plotted can be modified by whatplot argument first  predict value line second  confidence interval third  points fourth  axis fifth  summary of the end point sixth  corresponding line [CODESPLIT] nice.plot<-function(X=describe$generation,Y,maxY,minY=0,col,whatplot=c(T,T,T,T,T,T),ofset=0,name=\"\",labs=c(minY,ifelse(explarge==T,signif(maxY,nsignif),maxY)),span=0.18,end=length(X),roundrep=0,nsignif=3,percentual=F,explarge=F,labrep=ifelse(explarge==T,signif((ci[,1][newend]*(maxY-minY)+minY),nsignif),ifelse(percentual==T,paste(round((ci[,1][newend]*(maxY-minY)+minY)*100,roundrep),\"%\",sep=\"\"),round((ci[,1][newend]*(maxY-minY)+minY),roundrep))),cextext=1,lwdmain=2,mtextcex=1,firstax=-0.3,smooth=F){ col.li<-col col.ci<-paste(col,\"20\",sep=\"\") col.po<-paste(col,\"33\",sep=\"\") Y<-(Y-minY)/(maxY-minY) if(whatplot[3]==T){ points(X[1:end],Y[1:end],col=col.po,pch=16) } tryCatch({ lo <- loess(Y~X,degree=1,span=span) newX <- seq(min(X),max(X), length.out=1000) ci = cbind( predict(lo, data.frame(X=newX)), predict(lo, data.frame(X=newX))+ predict(lo, data.frame(X=newX), se=TRUE)$se.fit*qnorm(1-.05/2), predict(lo, data.frame(X=newX))- predict(lo, data.frame(X=newX), se=TRUE)$se.fit*qnorm(1-.05/2) ) }, error = function(e){ print(\"warning: loess returned error, original values instead of loess smoothed curved are visualized\") newX <- X ci = cbind(Y,Y,Y) }) newend<-which.min(abs(newX-end)) if(smooth==F){ newY <- approx(Y,n=1000)$y ci <- cbind(newY,newY,newY) } if(whatplot[1]==T){ lines(newX[1:newend], ci[,1][1:newend], lwd=lwdmain,col=col.li) } if(whatplot[2]==T){ if(smooth==T){ polygon(c(newX[1:newend],rev(newX[1:newend])), c(ci[,2][1:newend],rev(ci[,3][1:newend])), col=col.ci,border=NA) } } if(whatplot[4]==T){ par(mgp=c(3,ofset+firstax,ofset+firstax-0.2)) axis(2,at=c(0,1),lab=c(\"\",\"\"),col=col.li,col.axis=col.li,tck=-0.008) mtext(name,2,line=ofset+firstax,col=col.li,cex=mtextcex) mtext(labs,at=c(0,1),2,line=ofset+firstax,col=col.li,cex=mtextcex) } xtext<-par(\"usr\")[2]-(par(\"usr\")[2]-par(\"usr\")[1])/50 if(whatplot[5]==T){ text(xtext,ci[,1][newend],labrep,col=col.li,xpd=T,pos=4,cex=cextext) } if(whatplot[6]==T){ lines(c(0,xtext),rep(ci[,1][newend],2),col=col.li,lty=3,xpd=T) } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We draw a lollipop diagram to illustrate color distribution [CODESPLIT] par(ps = PS, mar = c(5, 5, 4, 2) + 0.1) plot(x = graphData$colour, y = graphData$Beach, type = \"p\", col = col_beach, ylim = c(0, 65), xlab = \"Colour\", ylab = \"Proportion [%]\", pch = PCH, xaxt = \"n\", cex = SS) axis(1, at = seq(1, 6, by = 1), labels = c(\"White\", \"Beige\", \"Grey\", \"Yellow\", \"Red\", \"Black\")) par(ps = PS, new = TRUE, mar = c(5, 5, 4, 2) + 0.1) plot(x = graphData$colour, y = graphData$Wreck, type = \"p\", col = col_wreck, ylim = c(0, 65), axes = FALSE, xlab = \"\", ylab = \"\", pch = PCH, cex = SS) arrows(x0 = graphData$colour, x1 = graphData$colour, y0 = graphData$Beach, y1 = graphData$Wreck, angle = 90, length = 0, ) legend(\"topright\", legend = c(\"Beach\", \"Wreck\"), fill = c(col_beach, col_wreck), bty = \"n\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Extract model statistics   Extract coefficients and std errors [CODESPLIT] model_coefs <- glmm_list %>% plyr::ldply(.fun = tidy) %>% subset(group == 'fixed') %>% droplevels", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Regression test for funnel plot asymmetry [CODESPLIT] res <- rma(Z_value,Z_var, data=data1, measure = \"ZCOR\", method = \"ML\") regtest(res, predictor = \"vi\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "build a new matrix that includes the formatted correlations and their significance stars [CODESPLIT] Rnew <- matrix(Rformatted, ncol = ncol(x)) rownames(Rnew) <- colnames(x) colnames(Rnew) <- paste(colnames(x), \"\", sep =\" \") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "CLEAR ENVIRONMENT   function to clear the environment after running a script while keeping important objects  use clear_environment: add store_to_keep variable with list variables;;  tip: include a variable with curr_env [CODESPLIT] clear_environment <- function(keep, # vector with variable-names to be excluded from cleaning all=ls(envir=pos.to.env(1))){ rm(list = all[! all %in% keep], envir = pos.to.env(1)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "boxplot visualizing the differences between the three languages [CODESPLIT] par(mfrow = c(1,1)) boxplot(ReadRate ~ Language, data = joint.read_rate, main = \"Reading rate, words/minute\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "zscore RSES T1 and T2 [CODESPLIT] df2$RSES1Z <- (df2$RSES1 - mean(df2$RSES1))/sd(df2$RSES1) df2$RSES2Z <- (df2$RSES2 - mean(df2$RSES2, na.rm = TRUE))/sd(df2$RSES2, na.rm = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " ANALYSES  Does Att Valence differ across Sections, OA Practice, and Level?  [CODESPLIT] mod_attvalence <- lmer(Att~Level*Section*OAPractice+(1|Responder_ID), df_attvalence) summary(mod_attvalence) anova(mod_attvalence) eta_sq(mod_attvalence, partial = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create variable Past_Future from Item. [CODESPLIT] df_use$Past_Future <- recode(df_use$Item, P1Use01 = \"In the past 12 months\",P1Use02 = \"In the following 12 months\", P2Use01 = \"In the past 12 months\",P2Use02 = \"In the following 12 months\", P3Use01 = \"In the past 12 months\",P3Use02 = \"In the following 12 months\", P4Use01 = \"In the past 12 months\",P4Use02 = \"In the following 12 months\", P5Use01 = \"In the past 12 months\",P5Use02 = \"In the following 12 months\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate average level of Att across Perspectives. [CODESPLIT] df_use_attvalence$Att <- rowMeans(df_use_attvalence[,c('Att_DailyLife', 'Att_ResearchField', 'Att_PublicSociety')], na.rm=TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Formula and family specification   In our case, setting `k  4` and `bs  'cr'` in the mgcv::s() smooth term is  not overly restrictive (checked by running the frequentist mgcv::gam()  function for each existing activity index in turn, thereby varying arguments  `k` and `bs`, and then checking the summary() of the resulting fits), but in a  future (prospective) study, the mgcv::s() settings need to be checked again:  [CODESPLIT] C_formula <- as.formula(paste(voutc, \"~\", paste0(\"s(\", vactidx_i, \", k = 4, bs = 'cr')\"))) C_family <- cumulative(link = \"logit\", link_disc = \"log\", threshold = \"flexible\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "display relationships between WVSES and associated variables [CODESPLIT] ggplot(df11, aes(WVSES, SE)) + geom_jitter() ggplot(df11, aes(WVSES, SISE)) + geom_jitter() ggplot(df11, aes(WVSES, ANX)) + geom_jitter() ggplot(df11, aes(WVSES, AVO)) + geom_jitter() ggplot(df11, aes(WVSES, LIFE)) + geom_jitter() ggplot(df11, aes(WVSES, EXTRA)) + geom_jitter() ggplot(df11, aes(WVSES, AGREE)) + geom_jitter() ggplot(df11, aes(WVSES, CON)) + geom_jitter() ggplot(df11, aes(WVSES, NEUR)) + geom_jitter() ggplot(df11, aes(WVSES, OPEN)) + geom_jitter() ggplot(df11, aes(WVSES, SSTA)) + geom_jitter() ggplot(df11, aes(WVSES, SINC)) + geom_jitter() ggplot(df11, aes(WVSES, MDS)) + geom_jitter() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "take RSES and run confirmatory factor analysis [CODESPLIT] RSESdata <- data.frame(df11$SE1, df11$SE2, df11$SE3, df11$SE4, df11$SE5, df11$SE6, df11$SE7, df11$SE8, df11$SE9, df11$SE10) singleF.model <- 'SE =~ df11.SE1 + df11.SE2 + df11.SE3 + df11.SE4 + df11.SE5 + df11.SE6 + df11.SE7 + df11.SE8 + df11.SE9 + df11.SE10' ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " pivot authorlevel data to \"longer\" authorarticlelevel dataframe  [CODESPLIT] article_author_level <- author_level %>% dplyr::select(FirstName_MI_LastName_str, full_name.kap, Article.1.Code.kap, Article.2.Code.kap, Article.3.Code.kap, Article.4.Code.kap, Article.5.Code.kap, Article.6.Code.kap, Article.7.Code.kap, Article.8.Code.kap, Article.9.Code.kap, Article.10.Code.kap, Gender.apsa, R_E.apsa, PhD.Year, PhD.Institution) %>% pivot_longer(cols = -c(FirstName_MI_LastName_str, full_name.kap, Gender.apsa, R_E.apsa, PhD.Year, PhD.Institution), values_to = \"article_title\", names_to = \"names\") %>% subset(article_title != \"NA\") %>% relocate(article_title) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "authorlevel count and proportion for all race / ethnic identity categories [CODESPLIT] summary_race_ethnicity_author <- article_author_level %>% dplyr::select(-c(article_title, names)) %>% distinct() %>% summarize(count = c(sum(white, na.rm = TRUE), sum(black, na.rm = TRUE), sum(east_asian, na.rm = TRUE), sum(south_asian, na.rm = TRUE), sum(latino, na.rm = TRUE), sum(mena, na.rm = TRUE), sum(native, na.rm = TRUE), sum(pacific, na.rm = TRUE), sum(other, na.rm = TRUE))) summary_race_ethnicity_author <- summary_race_ethnicity_author %>% mutate(race_ethnicity = c(\"White\", \"Black\", \"East Asian\", \"South Asian\", \"Hispanic or Latino\", \"Middle Eastern / Arab\", \"Native\", \"Pacific Islander\", \"Other\")) %>% mutate(proportion = round(count / sum(count), 2)) %>% dplyr::select(race_ethnicity, count, proportion) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Convert `candidate_page` to a factor. (Facilitates data visualizations and exploration). [CODESPLIT] elxn2015$candidate_page <- factor(elxn2015$candidate_page, levels = c(\"Harper\", \"Trudeau\", \"Mulcair\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Timestamps are initially encoded as factors. Convert them to POSIXct format to avoid errors. Dates and times are separated and saved as two new variables;; namely, `date_published` and `time_published`. `date_published` must be converted back to POSIXct format. `Time_published` is left as a character string as this format proves easier to work with when plotting timerelated data.  [CODESPLIT] elxn2015$created_time <- anytime(elxn2015$created_time, tz = \"America/Los_Angeles\") elxn2015$date_published <- as.POSIXct(format(elxn2015$created_time, \"%Y-%m-%d\")) elxn2015$time_published <- format(elxn2015$created_time, \"%H:%M:%S\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a new variable from `date_published` to indicate the month each post was published. [CODESPLIT] elxn2015$month_published <- as.POSIXct(elxn2015$date_published, format=\"%H:%M:%S\") elxn2015$month_published <- format(elxn2015$month_published, \"%B\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "`month_published` is converted to a factor with three levels such that values are chronologically ordered. [CODESPLIT] elxn2015$month_published <- factor(elxn2015$month_published, levels = c(\"August\", \"September\",\"October\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Ensure consistent data types across data sets. IDs are appropriately treated as characters rather than numeric. [CODESPLIT] Facebook_Pages$id <- as.character(Facebook_Pages$id) Facebook_Pages$likes_count <- as.double(Facebook_Pages$likes_count) Facebook_Pages$comments_count <- as.double(Facebook_Pages$comments_count) Facebook_Pages$shares_count <- as.double(Facebook_Pages$shares_count)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run random and fixed effects analyses on the data and calculate confidence intervals for the variables corresponding to heterogeneity (I2, H2 and T2). slab contains a vector of study names to display on the plot.  [CODESPLIT] res <- rma(measure = \"MD\", m1i = m2,m2i = m1, sd1i = sd2, sd2i = sd1, n1i = n2, n2i = n1, weights = as.numeric(weights.r), slab = c(\"Mc1\", \"Mc2\", \"Mi1\", \"Mi2\", \"Mi3\", \"Mi4\", \"Mi5\", \"Mi6\", \"Mi7\", \"Mi8\" ), method = \"DL\") res.f <- rma(measure = \"MD\", m1i = m2,m2i = m1, sd1i = sd2, sd2i = sd1, n1i = n2, n2i = n1, weights = weights.f, method = \"FE\") con <- confint(res) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "2. Transform precision into partial correlations for interpretation [CODESPLIT] pr2pc <- function(K) { D.Prec = diag(diag(K)^(-.5)) R <- diag(2,dim(K)[1])-D.Prec%*%K%*%D.Prec colnames(R) <- colnames(K) rownames(R) <- rownames(K) return(R) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "4. BDgraph extract posterior distribution for estimates [CODESPLIT] extractposterior <- function(fit, data, method = c(\"ggm\", \"gcgm\"), not.cont){ m <- length(fit$all_graphs) k <- 30000 n <- nrow(data) p <- ncol(data) j <- 1 densities <- rep(0, k) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "objects to store graph centrality measures [CODESPLIT] degree <- matrix(0, nrow = len, ncol = p) betweenness <- matrix(0, nrow = len, ncol = p) closeness <- matrix(0, nrow = len, ncol = p) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " create token named \"twitter_token\"  [CODESPLIT] twitter_token <- create_token( app = appname, consumer_key = key, consumer_secret = secret, access_token = access_token, access_secret = access_secret) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "mean/sd ratings of targets and decoys by condition [CODESPLIT] tapply(dat.long$rating, list(dat.long$condition, dat.long$target), mean) tapply(dat.long$rating, list(dat.long$condition, dat.long$target), function(x) sd(x)/sqrt(length(x))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "mean difference between candidates in each condition [CODESPLIT] round(s[s$target == 'envy.target' & s$condition == 'Baseline', \"emmean\"] - s[s$target == 'pity.target' & s$condition == 'Baseline', \"emmean\"],3) round(s[s$target == 'envy.target' & s$condition == 'PityAD', \"emmean\"] - s[s$target == 'pity.target' & s$condition == 'PityAD', \"emmean\"],3) round(s[s$target == 'envy.target' & s$condition == 'PityC', \"emmean\"] - s[s$target == 'pity.target' & s$condition == 'PityC', \"emmean\"],3) round(s[s$target == 'envy.target' & s$condition == 'EnvyAD', \"emmean\"] - s[s$target == 'pity.target' & s$condition == 'EnvyAD', \"emmean\"],3) round(s[s$target == 'envy.target' & s$condition == 'EnvyC', \"emmean\"] - s[s$target == 'pity.target' & s$condition == 'EnvyC', \"emmean\"],3) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "regression for difference between social/nonsocial [CODESPLIT] b = glmer(choice ~ pchoice1*social +  opchoice2*social + opchoice1 + pchoice2 + (1 + pchoice1 +  opchoice2 + opchoice1 + pchoice2  |code),family = 'binomial',data[data$opponent == 1,] ,control = glmerControl(optimizer=\"bobyqa\", tolPwrss = 1e-10, optCtrl = list(maxfun = 60000))) b = glmer(choice ~ pchoice1*social +  opchoice2*social + opchoice1 + pchoice2 + (1 + pchoice1 +  opchoice2 + opchoice1 + pchoice2  |code),family = 'binomial',data[data$opponent == 0,] ,control = glmerControl(optimizer=\"bobyqa\", tolPwrss = 1e-10, optCtrl = list(maxfun = 60000))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Principal component analysis with correlation matrix [CODESPLIT] pca <- prcomp(d[, -whi], scale. = T) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Make stacked figure of costs by age [CODESPLIT] stack <- lapply(pe.list.hadza_male, colMeans, na.rm=T) stacks <- as.data.frame(do.call(rbind, stack)) names(stacks) <- c(male.acts, \"climbing\") stacks$age <- age_seq stacks$Processing <- stacks$light_processing + stacks$other_out_of_camp + stacks$pounding_baobab stack_plot_male <- stacks %>% select(-nonsubsistence, -resting, -light_processing, -other_out_of_camp, -pounding_baobab) %>% gather(key=\"act\", value=\"cost\", carry_firewood_water:climbing, Processing) male_stack_hadza <- ggplot(stack_plot_male, aes(x=age, y=cost, fill=act)) + geom_area() + scale_fill_manual(name=\"Activity\", labels=c(\"Firewood and water collection\", \"Chopping\", \"Climbing\", \"Eating\", \"Manufacture\", \"Food processing\", \"Walking (during foraging)\"), values = brewer.pal(7, \"Set1\")) + labs(x=\"Age (years)\", y=\"Energy cost (kcal/day)\") + theme_classic(base_size=16) + ggtitle(\"Hadza: Men\") + lims(y=c(0, 750)) stack <- lapply(pe.list.hadza_female, colMeans, na.rm=T) stacks <- as.data.frame(do.call(rbind, stack)) names(stacks) <- female.acts stacks$age <- age_seq stacks$Processing <- stacks$light_processing + stacks$other_out_of_camp + stacks$pounding_baobab stack_plot_female <- stacks %>% select(-nonsubsistence, -resting, -light_processing, -other_out_of_camp, -pounding_baobab) %>% gather(key=\"act\", value=\"cost\", carry_firewood_water:digging, Processing) fhcols <- brewer.pal(8, \"Set1\")[c(1,2,4,5,6,7)] fhcols[2] <- \"#33CCFF\" female_stack_hadza <- ggplot(stack_plot_female, aes(x=age, y=cost, fill=act)) + geom_area() + scale_fill_manual(name=\"Activity\", labels=c(\"Firewood and water collection\", \"Digging\", \"Eating\", \"Manufacture\", \"Food processing\", \"Walking (during foraging)\"), values = fhcols) + labs(x=\"Age (years)\", y=\"Energy cost (kcal/day)\") + theme_classic(base_size=16) + ggtitle(\"Hadza: Women\") + lims(y=c(0, 750)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create centered age^2 variable for quadratic LGCM [CODESPLIT] qscale <- 100 rndhrs_subset <- rndhrs_subset %>% mutate(CAGE2_T4 = CAGE_T4^2 / qscale, CAGE2_T5 = CAGE_T5^2 / qscale, CAGE2_T6 = CAGE_T6^2 / qscale, CAGE2_T7 = CAGE_T7^2 / qscale, CAGE2_T8 = CAGE_T8^2 / qscale, CAGE2_T9 = CAGE_T9^2 / qscale, CAGE2_T10 = CAGE_T10^2 / qscale, CAGE2_T11 = CAGE_T11^2 / qscale ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "put.pvalue   Function to put the pvalues accoridng to the journals guidelines [CODESPLIT] put.pvalue=function(p){ sapply(p, function(x){ if (is.na(x)){pc=\"NA\"} else if (x<0.0001){pc=\"<0.0001\"} else if (0.0001<=x & x<0.001){pc <- sprintf(\"%.4f\", x)} else if (0.049<=x & x<0.05) {pc <- sprintf(\"%.3f\", floor(x*1000)/1000)} else if (0.001<=x & x<0.1){pc <- sprintf(\"%.3f\", x)} else if (0.1<=x){pc <- sprintf(\"%.2f\", x)} return(pc) }) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "wald.test   Function to perform Wald test in logistic or Cox model [CODESPLIT] wald.test <- function(model, ntest){ d <- length(ntest) a <- t(coefficients(model)[ntest]) mat <- vcov(model)[ntest,ntest] value <- as.numeric(a%*%solve(mat)%*%t(a)) pvalue <- pchisq(value, df=d, lower.tail=F) return(list(value=value, df=d, pvalue=pvalue)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function to change pvalues in the stat table with the ones from the model [CODESPLIT] update_table <- function(mymod) { modsum <- summary(mymod) p <- modsum$coefficients[, \"Pr(>|t|)\"] p <- as.vector(p) p <- pvalr(p, sig.limit = .001, digits = 3) p } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot KM Curve [CODESPLIT] tiff(\"PRAD_KM Curve.tiff\", units = \"in\", width=8, height=6, res=300) ggsurvplot(fit1, D2, risk.table=TRUE, palette = \"jco\", pval=TRUE, conf.int=FALSE, surv.median.line=\"hv\", ylab= \"Survival Probability (%)\", xlab= \"Time (months)\", legend.title = \"TX GROUP\", pval.coord=c(10,0.25)) dev.off() ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  now iterate through each participant, take that PP's data frame and add z scores for RT  afterwards, add that PP's data frame to the empty data frame to recreate the main df including z scores  [CODESPLIT] for (aPP in unique(excl2Exp$PP)){ dat <- excl2Exp[excl2Exp$PP == aPP,] dat$zScore <- scale(dat$RespRT, scale = TRUE) excl2Exp2 <- rbind(excl2Exp2, dat) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "ttest on subset 1 (n6/sex/group) [CODESPLIT] p_6m6f = t.test(measure ~ group, data=simdata6each_sub1)$p.value #using 6 each sex, t-tests ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "box and whisker plot for refined factor scores convert dataframe of factor scores and cluster membership to long form [CODESPLIT] KClusterScores.long <- KClusterScores %>% pivot_longer(c(Sensory, CognitiveDemand, ThreatToSelf, CrossSettings, Safety, States), names_to = c(\"Factor\")) KClusterScores.long$Factor <- factor(KClusterScores.long$Factor, levels = c(\"Sensory\", \"CognitiveDemand\", \"ThreatToSelf\", \"CrossSettings\", \"Safety\", \"States\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "box and whisker plot for nonrefined factor (scale) scores convert dataframe of factor scores and cluster membership to long form [CODESPLIT] scale.KClusterScores.long <- scale.KClusterScores %>% pivot_longer(c(Sensory, CognitiveDemand, ThreatToSelf, CrossSettings, Safety, States), names_to = c(\"Factor\")) scale.KClusterScores.long$Factor <- factor(scale.KClusterScores.long$Factor, levels = c(\"Sensory\", \"CognitiveDemand\", \"ThreatToSelf\", \"CrossSettings\", \"Safety\", \"States\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " calculate logrank Zstatistics assuming balanced risk set (in OE, E is always 0.5)  [CODESPLIT] zs <- sapply(1:(CureVobs1 + CureVobs0), function(n) (sum(CureVobs1s[1:n]) - n/2)/sqrt(n/4)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Statistics  check distributions  check normal distribution of PSE's [CODESPLIT] qqnorm(AllPSEs$Tennis.thre) shapiro.test(AllPSEs$Tennis.thre) # p < .01, qqnorm(AllPSEs$SoccerD.thre) shapiro.test(AllPSEs$SoccerD.thre) # p = .14, qqnorm(AllPSEs$SoccerI.thre) shapiro.test(AllPSEs$SoccerI.thre) # p < .01 qqnorm(AllPSEs$Coin.thre) shapiro.test(AllPSEs$Coin.thre) # p = .04 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "TEST 1: wilcoxon's signed rank test for direct/ indirect trials [CODESPLIT] (TestID <- wilcox.test(EffectDirect, EffectIndirect, paired = TRUE)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " add a column with \"time_interval\" : either 9:00;;13:00;;17:00;;21:00     for this, check the $created column and depending on which interval it is in, create a new column with mutate  [CODESPLIT] df_ESM_data_analysis = df_ESM_data_analysis %>% mutate( time_interval = case_when( format(created,\"%H:%M\") >= param_timepoints[1] & format(created,\"%H:%M\") < param_timepoints[2]  ~ param_timepoints[1], format(created,\"%H:%M\") >= param_timepoints[2] & format(created,\"%H:%M\") < param_timepoints[3]  ~ param_timepoints[2], format(created,\"%H:%M\") >= param_timepoints[3] & format(created,\"%H:%M\") < param_timepoints[4]  ~ param_timepoints[3], format(created,\"%H:%M\") >= param_timepoints[4] & format(created,\"%H:%M\") < param_t_end          ~ param_timepoints[4], ) ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "first find the duplicates according to n_obs and time_interval [CODESPLIT] df_ESM_data_analysis = df_ESM_data_analysis %>% arrange(session, created) %>% group_by(session, n_obs, time_interval) %>% mutate(num_dups = n(), dup_id = row_number()) %>% ungroup() %>% ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot perseverance as a function of n_obs for each subject [CODESPLIT] gdat <- groupedData(perseverance ~ n_obs | session, data=df_ESM_data_analysis_noNAnoOutliers) plot(gdat) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "put alpha together with descriptive statistics [CODESPLIT] questionnaire_stats <- bind_cols(questionnaires_stats, scores_alpha) %>% dplyr::select(-name, -vars) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Correlation plot (colors on top, numbers on bottom)  JUST MEANS (aggregated within person)  [CODESPLIT] corrplot(means, p.mat = res2$p, insig = \"label_sig\", sig.level = c(.001, .01, .05), pch.cex = .9, pch.col = \"gray\", method = \"color\", tl.col = \"black\", tl.cex = 1, cl.offset = 0.25, cl.cex = 0.75) corrplot(means, method=\"number\", col = \"black\", add = TRUE, cl.pos = \"n\", tl.pos = \"n\", type = \"lower\", number.cex = .60, number.digits = 2, tl.col = \"black\", cl.offset = 4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Independence test: two ways (HSIC or Nonlinear corr)  HSIC [CODESPLIT] if (independence == \"hsic\") { hsic.p <- dhsic.test(ry, rx, method = \"gamma\", kernel = \"gaussian\")$p.value pval[i] <- hsic.p } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Correlation table (only with complete cases  so takes out data from single parent) [CODESPLIT] means <- cor(mean_data_mat[,c(2:12)], use = \"complete.obs\") # Correlation matrix res2 <- cor.mtest(mean_data_mat[,c(2:12)], conf.level = .95, use = \"complete.obs\") # p-values ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "EXCLUDE PPN WITH RT < or > 3 SD [CODESPLIT] max_sd_rt <- print(mean(data.rt.check$rt)+3*sd(data.rt.check$rt)) ppn.exclude.rt <- print(filter(data.rt.check, rt >= max_sd_rt)$ppn) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate and plot difference between the distributions [CODESPLIT] postDiff <- forModel2plotA$b_cfaInequality_nat - forModel2plotB$b_cfaSocial_control_nat %>% as.data.frame() ggplot(postDiff, aes(.)) + geom_histogram(bins = 100, color = \"black\", fill = \"lightgray\") quantile(postDiff, c(0.025, 0.975)) #So 95% confidence interval of the difference lies above 0 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exclude participants with accuracy less than 0.8 [CODESPLIT] d_lowAcc =  CQ_acc %>% filter(accuracy < 0.8) d = d[!d$Subject %in% d_lowAcc$Subject,] rm(list=ls()[! ls() %in% c(\"d\")]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check the correlation between even and oddnumbered sets  Pearson [CODESPLIT] print(corr.test(person$ConditionOr_even, person$ConditionOr_odd), short = FALSE) KH_Correct(-.09) #.16 print(corr.test(person$ConditionOr_even, person$ConditionOr_odd), short = FALSE) KH_Correct(-.11) #.20 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "model criticism (remove data points with abs(standardized residuals) over 2.5) [CODESPLIT] de_trim<- de[abs(scale(resid(m_de)))<2.5,] do_trim<- do[abs(scale(resid(m_do)))<2.5,] m_de_mc = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = de_trim, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) m_do_mc = blmer(-1/Reaction.Time ~ Condition + (1+Condition|Subject) + (1+Condition|ItemNo), data = do_trim, control=lmerControl(optimizer = \"nloptwrap\", optCtrl = list(algorithm = \"NLOPT_LN_NELDERMEAD\", maxit = 2e5))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add new ParentIndex column in trialSummary and fill with NA [CODESPLIT] trialSummary[,\"ParentIndex\"] <- NA ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Lookup the parent row index of the stimulus string in s.data and insert it for social [CODESPLIT] for(k in 1:length(trialSummary$Stimulus)){ trialSummary$ParentIndex[k] <- which(s.data$Stimulus == trialSummary$Stimulus[k])[1] } trialSummary$AgeGroup <- s.info[i,3] trialSummary$Sex <- s.info[i,4] trialSummary$ID <- i ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fixation Count   custom mean function: mean of fixation count left and right eye (if only ONE eye contains 0.0 > no averaging)  https://stackoverflow.com/questions/45998419/aggregategroupmeanswhileignoringzerosunless0istheonlyvalue [CODESPLIT] meanSubset.FixCount = aggregate(Fixation.Count~Stimulus+AOI.Name, data=s.data,  FUN=(function(x){ifelse(sum(x==0)>0 & sum(x !=0) >0, mean(x[x>0]), mean(x))})) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "3) CALCULATION OF ESTIMATED MARGINAL MEANS   Tests for differences in response to jargon and get data for plots [CODESPLIT] at_jarg <- list(Jargon = c(\"Less\", \"More\"), TrvlAdv_Atten = \"Cons\", BackgrYrsOfExp = \"6-10 yrs\", BackgrDaysPerYr = \"11-20 days\", BullUseType = \"D\", cut = \"3|4\") grid_jarg <- ref_grid(clmm_underjar, mode = \"exc.prob\", at = at_jarg) (emm_jarg <- emmeans(grid_jarg, specs = pairwise ~ Jargon, by = \"BackgrAvTraining\")) plot_jarg <- summary(emm_jarg$emmeans) cont_jarg <- summary(emm_jarg$contrasts) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " create data frame for mean acceptrate (%) per participant, per effortlevel in 'young' group  [CODESPLIT] d.young.effort.1 <- filter(d, Group == 'Young') %>% group_by(ID, Agent, Effort) %>% summarise(m = mean(Choice)*100) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create data frame reflecting mean accept rate (%) per participant at each rewardlevel in 'old' group [CODESPLIT] d.reward.old.1 <- filter(d, Group == 'Old') %>% group_by(ID, Agent, Reward) %>% summarise(m = mean(Choice)*100) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create data frame reflecting overall mean accept rate (%) and SE at each rewardlevel in 'old' group [CODESPLIT] d.reward.old.2 <- d.reward.old.1 %>% group_by(Agent, Reward) %>% summarise(SE = std.error(m), M = mean(m, na.rm = T)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Welch's oneway ANOVA [CODESPLIT] anova_dehyd = df_dehyd %>% welch_anova_test(dehyd ~ cond) anova_dist_ran = df_dist_ran %>% welch_anova_test(dist ~ cond) anova_run_time = df_run_time %>% welch_anova_test(time ~ cond) anova_heat_time = df_heat_time %>% welch_anova_test(time ~ cond) anova_tcmax = df_tcmax %>% welch_anova_test(tcmax ~ cond) anova_tcmin = df_tcmin %>% welch_anova_test(tcmin ~ cond) anova_duration = df_duration %>% welch_anova_test(duration ~ cond) anova_thermal_load = df_thermal_load %>% welch_anova_test(thermal_load ~ cond) anova_heat_time = df_heat_time %>% welch_anova_test(heat_time ~ cond)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function to generate the true value for the Diamond Ratio [CODESPLIT] true.ratio <- function(n1, n2, tau, delta){ v <- (n1 + n2)/(n1*n2) + delta^2/2/(n1 + n2) j <- 1 - 3/(4*(n1 + n2 - 2) - 1) v.g <- (j^2)*v w <- 1/v.g w.star <- 1/(v.g + tau^2) var.fixed <- 1/sum(w) var.random <- 1/sum(w.star) sqrt(var.random/var.fixed) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "lmer model for dimorphism [CODESPLIT] Lat.ML<-lmer(Latency.cen~Sex+scale(Mass, scale = F)+ Generation+Pop+Stage+Injured+scale(Temp, scale = F)+ scale(Time, scale = F)+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(Lat.ML) drop1(Lat.ML) OF.ML<-lmer(sqrt(OF.Dist)~Sex+scale(Mass, scale = F)+ Generation+Pop+Stage+Injured+scale(Temp, scale = F)+ scale(Time, scale = F)+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(OF.ML) drop1(OF.ML) AP.ML<-lmer(sqrt(AP.Dist)~Sex+scale(Mass, scale = F)+ Generation+Pop+Stage+Injured+scale(Temp, scale = F)+ scale(Time, scale = F)+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(AP.ML) drop1(AP.ML) Lat.ML<-lmer(Latency.Cens~Sex+Mass+Generation+Pop+Stage+Injured+Temp+ Time+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(Lat.ML) drop1(Lat.ML) OF.ML<-lmer(OF.Dist~Sex+Mass+Generation+Pop+Stage+Injured+Temp+ Time+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(OF.ML) drop1(OF.ML) AP.ML<-lmer(AP.Dist~Sex+Mass+Generation+Pop+Stage+Injured+Temp+ Time+JDate_dev+(1|Sire)+(1|Dam),data=Data) summary(AP.ML) drop1(AP.ML) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "return the lastN characters of a string [CODESPLIT] fn$lastN<-function(string, n){ l<-nchar(string) result<-substr(string,l-n+1,l) result } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effects model for reproduction bias [CODESPLIT] mod_full <- lmer(Prct_Bias ~ ContextC*(GMSI_Gen_Z + OrderC*Distort) + (1|Subject), data=dat2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Extra function   MM to inchconversion for plot [CODESPLIT] MMtoInch <- function(MM) { z <- MM / 25.4 return(z) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot mean force (+/ SE) at each effortlevel in old group AND show individual data points [CODESPLIT] plot.force.old.2 <- ggplot(d.force.old, aes(x = Effort, y = m)) + geom_point(aes(group = Agent, color = Agent), position = position_dodge(width = 0.3), alpha = 0.4, shape = 1, size = 0.5) + geom_errorbar(data = d.force.old.summary, aes(group = Agent, color = Agent, ymin = m - SE, ymax = m + SE), position = position_dodge(0.02), alpha = 1, width = 0.2) + geom_line(data = d.force.old.summary, aes(group = Agent, color = Agent), position = position_dodge(0.02)) + scale_x_continuous(name = 'Effort level', breaks = seq(2, 6, 1)) + scale_y_continuous(name = 'Force exerted (normalised AUC)', breaks = seq(0.0, 1.0, 0.2)) + scale_color_brewer(palette = 'Set1') + coord_cartesian(xlim = c(2, 6), ylim = c(0.30, 1.0)) + theme_classic() + theme(legend.position = c(0.9, 0.15)) + theme(axis.text=element_text(size = 11), axis.title=element_text(size = 13)) plot.force.old.2 ggsave(filename = 'plot_force_old_v2.pdf', plot.force.old.2, height = 5, width = 4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot mean force (no SE) at each effortlevel in old group AND show individual data points [CODESPLIT] plot.force.old.3 <- ggplot(d.force.old, aes(x = Effort, y = m, group = Agent, color = Agent, fill = Agent)) + geom_point(data = d.force.old, aes(group = Agent, color = Agent), position = position_dodge(width = 0.3), alpha = 0.4, shape = 1, size = 0.5) + geom_smooth(method = 'lm', alpha = 0.3, size = 0.6, se = F) + coord_cartesian(xlim = c(2, 6), ylim = c(0.35, 0.9)) + scale_x_continuous(name = 'Effort level', breaks = seq(2, 6, 1)) + scale_y_continuous(name = 'Force exerted (normalised AUC)', breaks = seq(0, 1.0, 0.2)) + scale_color_brewer(palette = 'Set1') + scale_fill_brewer(palette = 'Set1') + theme_classic() + theme(legend.position = c(0.9, 0.15)) + theme(axis.text=element_text(size = 11), axis.title=element_text(size = 13)) plot.force.old.3 ggsave(filename = 'plot_force_old_v3.pdf', plot.force.old.3, height = 5, width = 4) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "bootstrap modindices of item residual covariances [CODESPLIT] cfa.csurf.bl.full.1.boot <- bootstrapLavaan(cfa.csurf.bl.full.1, R = 5000, FUN = function(x) {modindices(x)$mi}, verbose = TRUE) cfa.csurf.bl.full.1.boot <- data.frame(cfa.csurf.bl.full.1.boot) cfa.csurf.fu1.full.1.boot <- bootstrapLavaan(cfa.csurf.fu1.full.1, R = 5000, FUN = function(x) {modindices(x)$mi}, verbose = TRUE) cfa.csurf.fu1.full.1.boot <- data.frame(cfa.csurf.fu1.full.1.boot) cfa.csurf.bl.deponly.1.boot <- bootstrapLavaan(cfa.csurf.bl.deponly.1, R = 5000, FUN = function(x) {modindices(x)$mi}, verbose = TRUE) cfa.csurf.bl.deponly.1.boot <- data.frame(cfa.csurf.bl.deponly.1.boot) cfa.csurf.fu1.deponly.1.boot <- bootstrapLavaan(cfa.csurf.fu1.deponly.1, R = 5000, FUN = function(x) {modindices(x)$mi}, verbose = TRUE) cfa.csurf.fu1.deponly.1.boot <- data.frame(cfa.csurf.fu1.deponly.1.boot) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " recode played hours into numeric (if multiple numbers, choose the first one)  [CODESPLIT] mydf$Hours <- as.numeric(stri_extract_first_regex(mydf$Hours, \"[0-9]+\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " recode streamed hours into numeric (if multiple numbers, choose the first one)  [CODESPLIT] mydf$streams <- as.numeric(stri_extract_first_regex(mydf$streams, \"[0-9]+\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " recode narcisiim scale into numeric (if multiple numbers, choose the first one)  [CODESPLIT] mydf$Narcissism <- as.numeric(stri_extract_first_regex(mydf$Narcissism, \"[0-9]+\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Display and rank games by frequency [CODESPLIT] tGames <- as.data.frame(table(data_main$Game)) View(tGames[order(tGames$Freq, decreasing = TRUE), ]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Display and rank residence country by frequency [CODESPLIT] tResidence <- as.data.frame(table(data_main$Residence)) View(tResidence[order(tResidence$Freq, decreasing = TRUE), ]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Display and rank country of birth by frequency [CODESPLIT] tBirth <- as.data.frame(table(data_main$Birthplace)) View(tBirth[order(tBirth$Freq, decreasing = TRUE), ]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Loop to downsample data based on minimum time between points Time parameter is set in \"while(diff < time in minutes)\"  [CODESPLIT] trajectory_dsmpl <- data.frame() ids <- unique(trajectory.df$id) for(i in ids) { traj = subset(trajectory.df, trajectory.df$id == i) for(i in 1:nrow(traj)) { diff <- difftime(traj$dt[i+1], traj$dt[i], units = \"mins\") if(is.na(diff)) {break} while(diff < 15) { traj <- traj[-(i+1),] diff <- difftime(traj$dt[i+1], traj$dt[i], units = \"mins\") if(is.na(diff)) {break} } } trajectory_dsmpl <- bind_rows(trajectory_dsmpl, traj) } trajectory_dsmpl <- trajectory_dsmpl %>% group_by(id) %>% mutate(time_diff = difftime(dt, lag(dt, n = 1L), units = \"min\"), delta_x = x_utm - lag(x_utm, n = 1L), delta_y = y_utm - lag(y_utm, n = 1L), dist = sqrt(delta_x^2+delta_y^2)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Clean up dataframe and create a grouping variable 50m or 200m based on translocation distance [CODESPLIT] trajectory.df <- trajectory.df %>% ungroup %>% mutate(trans_group = ifelse(dist.y > 100, \"200m\", \"50m\")) %>% dplyr::select(id, sex, trans_group, trans_dist = dist.y, dt = dt.x, x_utm = x_utm.x, y_utm = y_utm.x, x_new, y_new, dist = dist.x, time_lag_min) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Barplot of homing success with percents as title 50m [CODESPLIT] bar_50m <- homing_variables %>% filter(trans_group == \"50m\") %>% ggplot(aes(x = sex, fill = sex, alpha = as.factor(homing_bin))) + geom_bar(position = \"fill\", color = \"black\", size = 2) + scale_fill_manual(values = c(\"#F8766D\", \"#00BFC4\")) + coord_flip() + theme(legend.position=\"none\", aspect.ratio = 0.4, axis.ticks.length = unit(.25, \"cm\"), axis.ticks.y = element_blank(), ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Barplot of homing success with percents as title 200m [CODESPLIT] bar_200m <- homing_variables %>% filter(trans_group == \"200m\") %>% ggplot(aes(x = sex, fill = sex, alpha = as.factor(homing_bin))) + geom_bar(position = \"fill\", color = \"black\", size = 2) + scale_fill_manual(values = c(\"#F8766D\", \"#00BFC4\")) + coord_flip() + theme(legend.position=\"none\", aspect.ratio = 0.4, axis.ticks.length = unit(.25, \"cm\"), axis.ticks.y = element_blank(), ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " 2 Factor Model, correlated factors  [CODESPLIT] model_2f <- 'f_a =~ parcel_a_1+parcel_a_2+parcel_a_3 # Agency f_c =~ parcel_c_1+parcel_c_2+parcel_c_3 # Communion ' fit.model_2f <- cfa(model_2f, data=data, std.lv=T) summary(fit.model_2f, fit.measures=TRUE, standardized = TRUE) # Robust model: CFI = .99, RMSEA = .02 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Factor Anaylsis with 3 factors [CODESPLIT] factors <- fa(items, nfactors = 3)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "binomial logistic regression testing for effect of condition [CODESPLIT] model1 <- glm(cbind(PredictTotal, PredictTrials-PredictTotal) ~ Condition, data = study1, family=binomial) summary(model1) Anova(model1, type=\"III\", test=\"Wald\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "binomial logistic regression testing for effects of behavior and context [CODESPLIT] model3 <- glm(cbind(PredictMatch, NumQs-PredictMatch) ~ Behavior + Context + Behavior*Context, data = study2, family=binomial) Anova(model3, type=\"III\", test=\"Wald\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function that creates a TRUE/FALSE matrix of nonbiological father availibility from two numbers that are recorded in the data  FROM and TILL. The function does this by evaluating the presence of the caretaker using < and >  for each interval in question. [CODESPLIT] giveage<-function(data){ FROM<-data$Grew_up_from_nonbiol TILL<-data$Grew_up_till_nonbiol ages<-data.frame(x=rep(NA,nrow(data))) for(i in 1:15){ j<-i-1 text<-paste(\"grew\",j,i,\"<-TILL>\",j,\"&FROM<\",i,sep=\"\") eval(parse(text=text)) text<-paste(\"ages<-cbind(ages,grew\",j,i,\")\",sep=\"\") eval(parse(text=text)) } ages<-ages[,-1] return(ages) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function that calculates differences between nonbiological fathers and partners in both groups (Nonbiological father currently present, nonbiological father currently absent) for each year in the analysis.  [CODESPLIT] deltas<-function(diff,ages){ deltaT<-NA deltaF<-NA for(i in 1:15){ deltaT[i]<-mean(diff[ages[i]==T],na.rm=T) deltaF[i]<-mean(diff[ages[i]==F],na.rm=T) } return(rbind(deltaT,deltaF)) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function that repeats last and first item of a vector  usefull if we want to draw plots and CI all the way to the border of the plotting region. [CODESPLIT] ad<-function(v){ return(c(v[1],v,v[length(v)])) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function that plots a text with an outline. it is equivalent to TeachingDemos shadowtext function described at: https://stackoverflow.com/questions/29303480/textlabelswithoutlineinr [CODESPLIT] shadowtext <- function(x, y=NULL, labels, col='white', bg='black', theta= seq(pi/4, 2*pi, length.out=40), r=0.1, ... ) { xy <- xy.coords(x,y) xo <- r*strwidth('A') yo <- r*strheight('A') for (i in theta) { text( xy$x + cos(i)*xo, xy$y + sin(i)*yo, labels, col=bg, ... ) } text(xy$x, xy$y, labels, col=col, ... ) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Functions coonducting logit and inverse logit transformation (p to logodds and back) [CODESPLIT] logit<-function(x){log(x/(1-x))} inv_logit<-function(x){exp(x)/(1+exp(x))}", "target": 1, "target_options": ["no_match", "match"]}
{"input": " read in all text files at working directory:  all_data  read_dir(      pattern  \"\\\\.txt$\",      stringsAsFactors  FALSE,      fill  TRUE,      header  TRUE  )  look at the columns and what they contain:  str(all_data)  look at data range, potential outliers:  peek_neat(all_data, 'rt')  the same per various groupings: ' peek_neat( '     dat  all_data, '     values  'rt', '     group_by  'response', '     round_to  1 ' ) ' peek_neat(all_data, 'rt', c('color', 'valence'), '           round_to  1)  histogram and QQ plots for the same:  peek_neat(all_data, 'rt', c('color', 'valence'), f_plot  plot_neat)  peek_neat(all_data, 'rt', c('color', 'valence'), f_plot  ggpubr::ggqqplot)  [CODESPLIT] filenames = list.files(pattern = \"^expsim_color_valence_.*\\\\.txt$\") # get all result file names for (file_name in enum(filenames)) { ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " look at rt data range and distribution, potential outliers  [CODESPLIT] peek_neat( data_final, values = c( 'rt_green_negative', 'rt_red_negative', 'rt_green_positive', 'rt_red_positive' ), group_by = 'condition', f_plot = plot_neat ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "now ANOVA on RTs for the main question: Color/Valence/Group interaction  with basic factorial plot of mean rt means  (95% CI for error bars by default) [CODESPLIT] anova_neat( data_final, values = c( 'rt_green_negative', 'rt_green_positive', 'rt_red_negative', 'rt_red_positive' ), within_ids = list( color = c('green', 'red'), valence = c('positive', 'negative') ), between_vars = 'condition', plot_means = TRUE, norm_tests = 'all', norm_plots = TRUE, var_tests = TRUE ) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "  kmeans clustering   Repeating exhaustive search for the kmeans clustering,  with a plot of the best clustering for each k  [CODESPLIT] KM2 <- kmeans.ex(PCA.Z$x, 2) KM2b <- KM2$best with(PCA.Z, { plot(x[, 1:2], pch = 20, asp = 1, col = cols[KM2b$cluster]) text(x[, 1:2], rownames(x), pos = 3) addhull(x[, 1], x[, 2], factor(KM2b$cluster), col.h = cols[1:2]) }) points(KM2b$centers[, 1], KM2b$centers[, 2], pch = 15, col = cols[1:2]) KM3 <- kmeans.ex(PCA.Z$x, 3) KM3b <- KM3$best with(PCA.Z, { plot(x[, 1:2], pch = 20, asp = 1, col = cols[KM3b$cluster]) text(x[, 1:2], rownames(x), pos = 3) addhull(x[, 1], x[, 2], factor(KM3b$cluster), col.h = cols[1:3]) }) points(KM3b$centers[, 1], KM3b$centers[, 2], pch = 15, col = cols[1:3]) KM4 <- kmeans.ex(PCA.Z$x, 4) KM4b <- KM4$best with(PCA.Z, { plot(x[, 1:2], pch = 20, asp = 1, col = cols[KM4b$cluster]) text(x[, 1:2], rownames(x), pos = 3) addhull(x[, 1], x[, 2], factor(KM4b$cluster), col.h = cols[1:4]) }) points(KM4b$centers[, 1], KM4b$centers[, 2], pch = 15, col = cols[1:4]) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "write the number of observed variables to the top [CODESPLIT] header[1] <- paste(ifelse(Flow==FALSE , length(unique(df$Point)), length(unique(df$Point)) + 1), \"     : number of observed variables\") write(header[1:(grep(\"subbasin number\",header) - 2)], file = outfile) header[1] <- paste(n + n1,\"     : number of observed variables\") write(header[1:(grep(\"subbasin number\",header) - 2)], file = outfile) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " \"Remove\" first word of each page, except on the first page where the title was the first word (but the title has already been removed)   [CODESPLIT] PAST_M_21$gazedur[!duplicated(PAST_M_21$page) & PAST_M_21$page > 1] <- NaN PAST_M_21$fixdur[!duplicated(PAST_M_21$page) & PAST_M_21$page > 1] <- NaN PRES_O_21$gazedur[!duplicated(PRES_O_21$page) & PRES_O_21$page > 1] <- NaN PRES_O_21$fixdur[!duplicated(PRES_O_21$page) & PRES_O_21$page > 1] <- NaN PRES_M_21$gazedur[!duplicated(PRES_M_21$page) & PRES_M_21$page > 1] <- NaN PRES_M_21$fixdur[!duplicated(PRES_M_21$page) & PRES_M_21$page > 1] <- NaN PAST_O_21$gazedur[!duplicated(PAST_O_21$page) & PAST_O_21$page > 1] <- NaN PAST_O_21$fixdur[!duplicated(PAST_O_21$page) & PAST_O_21$page > 1] <- NaN", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a vector with the rounded values (names(valM)  adj_dimension) [CODESPLIT] valM <- round(colMeans(d, na.rm = T), 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Divide estimates, posterior sd, lower CI, and upper CI of the withinperson effects by the withinperson SDs of social interactions  to obtain coefficients that are standardized with respect to the DV only  [CODESPLIT] Values_Analysis1_Model1$est[rows] <- Values_Analysis1_Model1$est[rows] / sqrt(variances) Values_Analysis1_Model1$posterior_sd[rows] <- Values_Analysis1_Model1$posterior_sd[rows] / sqrt(variances) Values_Analysis1_Model1$lower_2.5ci[rows] <- Values_Analysis1_Model1$lower_2.5ci[rows] / sqrt(variances) Values_Analysis1_Model1$upper_2.5ci[rows] <- Values_Analysis1_Model1$upper_2.5ci[rows] / sqrt(variances) Values_Analysis1_Model1_Pers$est[rows] <- Values_Analysis1_Model1_Pers$est[rows] / sqrt(variances) Values_Analysis1_Model1_Pers$posterior_sd[rows] <- Values_Analysis1_Model1_Pers$posterior_sd[rows] / sqrt(variances) Values_Analysis1_Model1_Pers$lower_2.5ci[rows] <- Values_Analysis1_Model1_Pers$lower_2.5ci[rows] / sqrt(variances) Values_Analysis1_Model1_Pers$upper_2.5ci[rows] <- Values_Analysis1_Model1_Pers$upper_2.5ci[rows] / sqrt(variances)", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Heart rate  median considered as mean when mean not \"available\"  [CODESPLIT] DATA$HR <-  ifelse(is.na(DATA$\"all_mean_HR\")==T, DATA$all_median_HR, DATA$\"all_mean_HR\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Keep the removed people to plot separately [CODESPLIT] nograph_removed <- nograph[nograph$belief_in_medicine >= 7,] graph_removed <- graph[graph$belief_in_medicine >= 7,] ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Means and Standard Deviations of Conditions  Experiencer [CODESPLIT] exp <- data.all[ which(data.all$condition == 'experiencer'),] round(mean(exp$rating), 2) round(sd(exp$rating), 2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Q3 create a new data set that contains the first 300 cases from the subset you have just created above [CODESPLIT] working_data_2 <- slice(working_data_1, 1:300) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Q5 create one single subset (based on your original dataset) where you   select the variables id sex age source1 discuss flushot vacc1 and refus   select cases 150  450 only   change the name of the variable source1 to 'Main_Source'   select particpants older than 39 years of age, and who vaccinate their own children (mandatory and reccommended vaccines) hint: %>%  MAGRITTR package  [CODESPLIT] working_data_final <- ex1_data %>% select(id, sex, age, source1, discuss, flushot, vacc1, refus) %>% rename(Main_Source=source1) %>% slice(150:450) %>% filter(age > 39 & vacc1 == 'Mandatory + all recommended') ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " calculating r for each iteration, r as the mean of those five r  RAQ_R and QoL  [CODESPLIT] r_QoL <- (with(subsample_PG, cor(RAQ_Totalscore[.imp==1 & Sample==1], QoL[.imp==1 & Sample==1]))+ with(subsample_PG, cor(RAQ_Totalscore[.imp==2 & Sample==1], QoL[.imp==2 & Sample==1]))+ with(subsample_PG, cor(RAQ_Totalscore[.imp==3 & Sample==1], QoL[.imp==3 & Sample==1]))+ with(subsample_PG, cor(RAQ_Totalscore[.imp==4 & Sample==1], QoL[.imp==4 & Sample==1]))+ with(subsample_PG, cor(RAQ_Totalscore[.imp==5 & Sample==1], QoL[.imp==5 & Sample==1])))/5 p_QoL <- pt((r_QoL*sqrt(nrow(subsample_PG)/6-2)/sqrt(1-r_QoL)), nrow(subsample_PG)/6-2) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "scaling RAQR Score [CODESPLIT] reg_with_PG_CG_scaled <- with(data=as.mids(reg_PG_CG), exp=lm(scale(RAQ_Totalscore)~Sample+Gender+Age_groups+Level_of_education)) str(summary(pool(reg_with_PG_CG_scaled))) pooled_reg_PG_CG_scaled <- summary(pool(reg_with_PG_CG_scaled)) pooled_reg_PG_CG_scaled$p.value pooled_reg_PG_CG_scaled$estimate ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "regression for each iteration in order to check diagnostic plots  look for differences between the adjusted Rsquared of the five iterations [CODESPLIT] reg_PG_CG_1 <- reg_PG_CG[reg_PG_CG$.imp == 1, ] summary(lm(data=reg_PG_CG_1, RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) reg_with_PG_CG_1 <- with(data=reg_PG_CG_1, exp=lm(RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) plot(reg_with_PG_CG_1) reg_PG_CG_2 <- reg_PG_CG[reg_PG_CG$.imp == 2, ] summary(lm(data=reg_PG_CG_2, RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) reg_with_PG_CG_2 <- with(data=reg_PG_CG_2, exp=lm(RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) plot(reg_with_PG_CG_2) reg_PG_CG_3 <- reg_PG_CG[reg_PG_CG$.imp == 3, ] summary(lm(data=reg_PG_CG_3, RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) reg_with_PG_CG_3 <- with(data=reg_PG_CG_3, exp=lm(RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) plot(reg_with_PG_CG_3) reg_PG_CG_4 <- reg_PG_CG[reg_PG_CG$.imp == 4, ] summary(lm(data=reg_PG_CG_4, RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) reg_with_PG_CG_4 <- with(data=reg_PG_CG_4, exp=lm(RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) plot(reg_with_PG_CG_4) reg_PG_CG_5 <- reg_PG_CG[reg_PG_CG$.imp == 5, ] summary(lm(data=reg_PG_CG_5, RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) reg_with_PG_CG_5 <- with(data=reg_PG_CG_5, exp=lm(RAQ_Totalscore~Sample+Gender+Age_groups+Level_of_education)) plot(reg_with_PG_CG_5) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " calculating the mean of the five residual standard errors, R squared, adjusted Rsquared and F statistic  [CODESPLIT] Residual_standard_error_reg_PG <- (10.35+10.45+10.16+10.29+10.28)/5 R_squared_reg_PG <- (0.639+0.630+0.652+0.642+0.642)/5 R_squared_adjusted_reg_PG <-(0.6267+0.6175+0.6403+0.6301+0.6298)/5 F_statistic_reg_PG <- (53.04+51.05+56.18+53.81+53.73)/5", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Graph the correlation between secure attachment and visual cortex response [CODESPLIT] ggplot(mydata, aes(SECURE.ATTACHMENT, LINGUAL.GYRUS..VISUAL.CORTEX.), scale=\"globalminmax\") + geom_smooth(method = \"lm\", fill = \"green\", alpha = 0.6)+ geom_point(size =5 ) + theme_minimal()+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(),panel.grid.minor = element_blank(), axis.line = element_line(colour = \"black\"))+ labs(title = \"Visual Cortex & Secure Attachment\", x = \"Secure Attachment\", y = \"Lingual Gyrus Activation\") scale_x_continuous(limits = c(1,5), breaks = c(1,2,3,4,5)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Graph the correlation between avoidant attachment and visual cortex response [CODESPLIT] ggplot(mydata, aes(AVOIDANT.ATTACHMENT, LINGUAL.GYRUS..VISUAL.CORTEX.), scale=\"globalminmax\") + geom_smooth(method = \"lm\", fill = \"red\")+ geom_point(size = 5) + theme_minimal()+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(),panel.grid.minor = element_blank(), axis.line = element_line(colour = \"black\"))+ labs(title = \"Visual Cortex & Avoidant Attachment\", x = \"Avoidant Attachment\", y = \"Lingual Gyrus Activation\")+ scale_x_continuous(limits = c(1,5), breaks = c(1,2,3,4,5)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Graph the interaction plot of amygdala and visual cortex activation, with AVOIDANT attachment as the moderator.  [CODESPLIT] p1 = interact_plot(fiti, pred = AMYGDALA, modx = AVOIDANT.ATTACHMENT,robust = FALSE, x.label = \"Amygdala\", y.label = \"LG Visual Cortex\", main.title = \"Avoidant Attachment\",  legend.main = \"Avoidant Levels\", colors = \"red\",interval = TRUE, int.width = 0.8)+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(),panel.grid.minor = element_blank(), axis.line = element_line(colour = \"black\")) p1 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Graph the interaction plot of amygdala and visual cortex activation, with SECURE attachment as the moderator.  [CODESPLIT] p2 = interact_plot(fiti2, pred = AMYGDALA, modx = SECURE.ATTACHMENT,robust = FALSE, x.label = \"Amygdala\", y.label = \"LG Visual Cortex\", main.title = \"Secure Attachment\",  legend.main = \"Secure Levels\", colors = \"green\",interval = TRUE, int.width = 0.8)+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(),panel.grid.minor = element_blank(), axis.line = element_line(colour = \"black\")) p2 ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Returns the ratio between the geometric means of two independent samples and its  95% BCa bootstrap CI. [CODESPLIT] ratioGeomMeanCI.bootstrap <- function(group1, group2) { group1 <- log(group1) group2 <- log(group2) samplemean <- function(x, d) {return(mean(x[d]))} pointEstimate <- samplemean(group1) - samplemean(group2) set.seed(0) # make deterministic bootstrap_samples <- two.boot(sample1 = group1, sample2 = group2, FUN = samplemean, R = 5000) bootci <- boot.ci(bootstrap_samples, type = \"bca\", conf = conf.level) exp(c(pointEstimate,  bootci$bca[4], bootci$bca[5])) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Returns the 95% confidence interval of a single proportion using the Wilson score interval. [CODESPLIT] propCI <- function(numberOfSuccesses, sampleSize) { CI <- scoreci(x = numberOfSuccesses, n = sampleSize, conf.level = conf.level) c(numberOfSuccesses/sampleSize, CI$conf.int[1], CI$conf.int[2]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Returns the difference between two linear regression slopes and its 95% BCa bootstrap CI. [CODESPLIT] diff.slopes.bootstrap <- function(x1, y1, x2, y2) { groups <- c(rep(1, length(x1)), rep(2, length(y1)), rep(3, length(x2)), rep(4, length(y2))) data <- data.frame(obs = c(x1, y1, x2, y2), group = groups) diffslope <- function(d, i) { db <- d[i,] x1 <- db[db$group==1,]$obs y1 <- db[db$group==2,]$obs x2 <- db[db$group==3,]$obs y2 <- db[db$group==4,]$obs fit1 <- lm(y1 ~ x1) a1 <- fit1$coefficients[[2]] fit2 <- lm(y2 ~ x2) a2 <- fit2$coefficients[[2]] a2 - a1 } a1 <- lm(y1 ~ x1)$coefficients[[2]] a2 <- lm(y2 ~ x2)$coefficients[[2]] pointEstimate <- a2 - a1 set.seed(0) # make deterministic bootstrap_samples <- boot(data = data, statistic = diffslope, stype = \"i\", strata = data$group, R = 5000) bootci <- boot.ci(bootstrap_samples, type = \"bca\") c(pointEstimate, bootci$bca[4], bootci$bca[5]) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "replace current subjects \"fixdata\" by fixation_info  [CODESPLIT] if (any(S==replace_eye)){ fixdata[[S]] <- data.frame(fixation_info) colnames(fixdata[[S]]) <- c(\"Event.Start.Raw.Time..ms.\",\"Event.End.Raw.Time..ms.\",\"Event.Duration.Trial.Time..ms.\", \"Fixation.Position.X..px.\",\"Fixation.Position.Y..px.\",\"AOI.Name\",\"intrialonset\",\"trialnr\") } } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mean Differences Comparing signed and unsigned reviews on four aspects For each, first ttest then violin plot Paneled plot created at end of series Testing word count  [CODESPLIT] t.test(revdat$wc ~ revdat$signed) wc_sign <- ggplot(revdat, aes(signed, wc, fill=signed)) + geom_violin( trim = FALSE, draw_quantiles = c(0.25, 0.5, 0.75), alpha = 0.5) + geom_jitter( width = 0.20, height = 0, alpha = 0.5, size = 1) + xlab(\"Signed Reviews\") + ylab(\"Word Count\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Correlations Examining correlations between time and aspects of reviews For each, first compute r and then scatter plot Plots also mark whether or not review was signed Paneled plot created at end of series  [CODESPLIT] cor.test(revdat$order, revdat$wc) wc_time <- ggplot(revdat, aes(order, wc, color=signed)) + geom_point(size=1)+ xlab(\"Time\") + ylab(\"Word Count\") cor.test(revdat$order, revdat$posemo) pos_time <- ggplot(revdat, aes(order, posemo, color=signed)) + geom_point(size=1) + xlab(\"Time\") + ylab(\"Positive Emotion Words\") cor.test(revdat$order, revdat$negemo) neg_time <- ggplot(revdat, aes(order, negemo, color=signed)) + geom_point(size=1) + xlab(\"Time\") + ylab(\"Negative Emotion Words\") cor.test(revdat$order, revdat$cogmech) cog_time <- ggplot(revdat, aes(order, cogmech, color=signed)) + geom_point(size=1) + xlab(\"Time\") + ylab(\"Cognitive Mechanism Words\") ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "removes nonalphanumeric characters [CODESPLIT] full.data <- multigsub(\"[^[:alnum]]\", \" \", full.data, fixed = TRUE) train.data <- multigsub(\"[^[:alnum]]\", \" \", train.data, fixed = TRUE) valid.data <- multigsub(\"[^[:alnum]]\", \" \", valid.data, fixed = TRUE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "removes leading & trailing whitespaces [CODESPLIT] full.data <- trimws(full.data) train.data <- trimws(train.data) valid.data <- trimws(valid.data)", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal glmer() model  This model contains a fixed withinsubjects effect of Ambiguity (effectcoded with 0.5 amb),  codes for modality effects and interactions,  plus random effects by participants and items.  [CODESPLIT] Acc.max <- glmer(Correct ~ 1 + Ambiguity.code + Modality.code1 + Modality.code2 + Interaction.code1 + Interaction.code2 + (1 + Ambiguity.code + Modality.code1 + Modality.code2 + Interaction.code1 + Interaction.code2 | Participant.Private.ID) + (1   | Item), data = Data.CohOnly, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.max <- glmer(Correct ~ 1 + Ambiguity.code + Modality.code2 + Interaction.code2 + (1 + Ambiguity.code + Modality.code2 + Interaction.code2 | Participant.Private.ID) + (1   | Item), data = Data.ListandRead, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) Acc.max <- glmer(Correct ~ 1 + Ambiguity.code + Modality.code3 + Interaction.code3 + (1 + Ambiguity.code + Modality.code3 + Interaction.code3 | Participant.Private.ID) + (1   | Item), data = Data.ListandRSVP, family = \"binomial\", control = glmerControl(optimizer =\"bobyqa\")) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal lmer() model  This model contains codes for modality effects,  plus random effects by participants and items.  [CODESPLIT] RT.AmbOnly.max <- lmer(logRT ~ 1 + Modality.code1 + Modality.code2 + (1 + Modality.code1 + Modality.code2 | Participant.Private.ID) + (1   | Item), data = Data.AmbOnly, REML=FALSE) RT.ListandRead.max <- lmer(logRT ~ 1 + Modality.code2 + (1 + Modality.code2 | Participant.Private.ID) + (1   | Item), data = Data.ListandRead, REML=FALSE) RT.ListandRSVP.max <- lmer(logRT ~ 1 + Modality.code3 + (1 + Modality.code3 | Participant.Private.ID) + (1   | Item), data = Data.ListandRSVP, REML=FALSE) RT.ReadandRSVP.max <- lmer(logRT ~ 1 + Modality.code3 + (1 + Modality.code3 | Participant.Private.ID) + (1   | Item), data = Data.ReadandRSVP, REML=FALSE) RT.UAOnly.max <- lmer(logRT ~ 1 + Modality.code1 + Modality.code2 + (1 + Modality.code1 + Modality.code2 | Participant.Private.ID) + (1   | Item), data = Data.UAOnly, REML=FALSE) RT.ListandRead.max <- lmer(logRT ~ 1 + Modality.code2 + (1 + Modality.code2 | Participant.Private.ID) + (1   | Item), data = Data.ListandRead, REML=FALSE) RT.ListandRSVP.max <- lmer(logRT ~ 1 + Modality.code3 + (1 + Modality.code3 | Participant.Private.ID) + (1   | Item), data = Data.ListandRSVP, REML=FALSE) RT.ReadandRSVP.max <- lmer(logRT ~ 1 + Modality.code3 + (1 + Modality.code3 | Participant.Private.ID) + (1   | Item), data = Data.ReadandRSVP, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " construct a maximal lmer() model  This model contains a fixed effect for Ambiguity,  plus random effects by participants and items.  [CODESPLIT] RT.ListeningOnly.max <- lmer(logRT ~ 1 + Ambiguity.code + (1 + Ambiguity.code | Participant.Private.ID) + (1   | Item), data = Data.ListeningOnly, REML=FALSE) RT.ReadingOnly.max <- lmer(logRT ~ 1 + Ambiguity.code + (1 + Ambiguity.code | Participant.Private.ID) + (1   | Item), data = Data.ReadingOnly, REML=FALSE) RT.RSVPOnly.max <- lmer(logRT ~ 1 + Ambiguity.code + (1 + Ambiguity.code | Participant.Private.ID) + (1   | Item), data = Data.RSVPOnly, REML=FALSE) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " As this may be problematic for later analyses, we transform urbanity into its log:  [CODESPLIT] dslong %<>% mutate(urbanity_log = log(urbanity))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate Bayes factors for difference using logspline fit [CODESPLIT] prior <- dnorm(0,1) fit.posterior <- logspline(samples$BUGSoutput$sims.list$mu_alpha) posterior <- dlogspline(0, fit.posterior) # this gives the pdf at point delta = 0 prior/posterior ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "use normal distribution to approximate pvalue [CODESPLIT] FullAim1aPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim1aPNcoefs$t.value))) FullAim1aPNcoefs effectsize::standardize_parameters(FullAim1aPN) FullAim1bPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim1bPNcoefs$t.value))) FullAim1bPNcoefs effectsize::standardize_parameters(FullAim1bPN) FullAim1cPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim1cPNcoefs$t.value))) FullAim1cPNcoefs effectsize::standardize_parameters(FullAim1cPN) FullAim2aPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim2aPNcoefs$t.value))) FullAim2aPNcoefs effectsize::standardize_parameters(FullAim2aPN) FullAim2bPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim2bPNcoefs$t.value))) FullAim2bPNcoefs effectsize::standardize_parameters(FullAim2bPN) FullAim3PNPPcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim3PNPPcoefs$t.value))) FullAim3PNPPcoefs effectsize::standardize_parameters(FullAim3PNPP) FullAim4aPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim4aPNcoefs$t.value))) FullAim4aPNcoefs effectsize::standardize_parameters(FullAim4aPN) FullAim4bPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim4bPNcoefs$t.value))) FullAim4bPNcoefs effectsize::standardize_parameters(FullAim4bPN) FullAim5bPosPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim5bPosPNcoefs$t.value))) FullAim5bPosPNcoefs effectsize::standardize_parameters(FullAim5bPosPN) FullAim5bNegPNcoefs$p.z <- 2 * (1 - pnorm(abs(FullAim5bNegPNcoefs$t.value))) FullAim5bNegPNcoefs effectsize::standardize_parameters(FullAim5bNegPN)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "this function give the integral of the survival curve given by S.hat ont the time.grid: Y.grid [CODESPLIT] expected_survival <- function(S.hat, Y.grid) { grid.diff <- diff(c(0, Y.grid, max(Y.grid))) c(base::cbind(1, S.hat) %*% grid.diff) } threshol_list <- function(l,threshold){ l_threshold = c() for (x in (l)){ if (is.na(x)){ return(NA) } else{ if (x < - threshold){ l_threshold <- c(l_threshold,- threshold) } else{ if(x > threshold){ l_threshold <- c(l_threshold,threshold) } else{ l_threshold <- c(l_threshold, x) } } } } return(l_threshold) } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "change data type of PHDYEAR to numeric [CODESPLIT] author_phd_data$PHDYEAR <-as.numeric(author_phd_data$PHDYEAR)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate seniority of each author at the time of each article [CODESPLIT] author_phd_data$status_article1 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.1.Year.published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article1 == 0|author_phd_data$yrs_to_article1 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article1 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article2 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.2.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article2 == 0|author_phd_data$yrs_to_article2 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article2 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article3 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.3.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article3 == 0|author_phd_data$yrs_to_article3 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article3 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article4 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.4.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article4 == 0|author_phd_data$yrs_to_article4 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article4 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article5 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.5.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article5 == 0|author_phd_data$yrs_to_article5 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article5 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article6 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.6.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article6 == 0|author_phd_data$yrs_to_article6 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article6 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article7 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.7.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article7 == 0|author_phd_data$yrs_to_article7 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article7 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article8 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.8.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article8 == 0|author_phd_data$yrs_to_article8 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article8 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article9 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.9.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article9 == 0|author_phd_data$yrs_to_article9 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article9 > 6, \"Senior Scholar\", \"NA\"))) author_phd_data$status_article10 <- ifelse(author_phd_data$PHDYEAR == 0 |author_phd_data$PHDYEAR > author_phd_data$Article.10.Year.Published, \"Grad Student\", ifelse(author_phd_data$yrs_to_article10 == 0|author_phd_data$yrs_to_article10 < 7, \"Junior Scholar\", ifelse(author_phd_data$yrs_to_article10 > 6, \"Senior Scholar\", \"NA\"))) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run check to see if any authors have missing data [CODESPLIT] check <- subset(mmcpsr_authors, is.na(mmcpsr_authors$Title)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "stacked graph regarding how participants came to see each advocacy type [CODESPLIT] ggplot(detected_advocacy, aes(x = Advocacy, y = Percentage, fill = Response, label = Percentage)) + geom_bar(position =\"stack\", stat=\"identity\") + coord_flip(ylim=c(0,100)) + scale_y_continuous(labels = scales::percent_format(scale = 1)) + geom_text(aes(label = Percentage), size = 3, position = position_stack(vjust = 0.5) ) + labs(y= \"Porcentaje de Participantes\", x = \"Tipo de Defensa\") + theme(text = element_text(family = \"Arial\", size = 14), panel.background = element_rect(\"white\"), panel.border = element_rect(fill = NA), panel.grid.major.x = element_line(colour = \"grey\"), axis.title.x = element_text(color = \"#39363D\", size = 14, hjust=0.5), axis.title.y = element_text(color = \"#39363D\", size = 14), legend.position=\"bottom\", legend.text = element_text(size = 12), legend.title = element_blank()) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Define adjustment to calculate Hedge's g. To calculate Cohen's d set J < 1. Remember to change the true.ratio value of J as well. [CODESPLIT] J <-   j <- 1 - 3/(4*(n + m - 2) - 1) G <- J*SMD sds <- sqrt((n + m)/(n*m) + SMD^2/2/(n + m)) V <- (J^2)*(sds^2) w <- 1/V if(method == \"REML\"){r.model <- rma(yi = G, vi = V,method = method, control=list(stepadj=0.5, maxiter=10000000000000000000000000))} else{ r.model <- rma(yi = G, vi = V,method = method) } fit.model <- rma(yi = G, vi = V, method = \"FE\") ratio <- abs(r.model$ci.ub - r.model$ci.lb)/abs(fit.model$ci.ub - fit.model$ci.lb) log.ratio <- log(ratio) w.star <- 1/(V + r.model$tau2) log.var <- ((r.model$se.tau2)^2)/4*1/(sum(w.star))^2*(sum(w.star^2))^2 log.sd <- sqrt(log.var) bias <- 1/2*(r.model$se.tau2)^2*(1/2/sum(w.star)^2 - 1/sum(w.star)*sum(w.star^3)) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate parcellated data and do CFA   generate parcellated datasets  must specify data and the number of allocations (nAlloc) [CODESPLIT] list1=parcelAllocation(mod.par, data=usm, par.names, mod.items, nAlloc=2, do.fit=F, std.lv=T) ", "target": 1, "target_options": ["no_match", "match"]}
{"input": " Multifit() conducts a CFA for each data.frame in a list (saved from parcelAllocation) and returns all of the results of interest on one row per data.frame  The results consist of 21 columns with fit measures (\"npar\", \"chisq\", \"df\", \"pvalue\", \"cfi\", \"tli\", \"rmsea\", \"rmsea.pvalue\", \"srmr\") followed by the latent r between scales (\"IA_NFC\", \"IA_URS\", etc.)  WARNING: running each CFA takes roughly 20 seconds (Intel i58350U CPU), so running it for 200 data.frames should take 11.5 hours.  [CODESPLIT] multifit=function(data) { rows=c(131:136) # rows indexing the parameter estimates of interest (correlations among general factors) names=c(\"npar\", \"chisq\", \"df\", \"pvalue\", \"cfi\", \"tli\", \"rmsea\", \"rmsea.pvalue\", \"srmr\", \"IA_NFC\", \"IA_IU\", \"IA_URS\", \"NFC_IU\", \"NFC_URS\", \"IU_URS\", \"IA_NFC_se\", \"IA_IU_se\", \"IA_URS_se\", \"NFC_IU_se\", \"NFC_URS_se\", \"IU_URS_se\") results=sapply(data, function(x){ output=cfa(data=x, model=mod.par, std.lv=T) # save CFA output est=as.matrix(out@ParTable[[\"est\"]][rows]) # save the parameter estimates of interest se=as.matrix(out@ParTable[[\"se\"]][rows]) fit=as.matrix(fitMeasures(output)[c(1,3,4,5,9,10,23,26,29)]) # save fit measures of interest bind=round(do.call(\"rbind\", (list(fit,est,se))),3) # bind the data }) results.t=t(results) # transpose the matrix colnames(results.t)=names # name columns results.t } ", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Get residuals of tie strength adjusted tie strength model_tiestrength glm(strength.mean~no.of.papers:Gender, family\"gaussian\", data new) model with strength not transformed looked a bit dodgy, so we logtransformed strength [CODESPLIT] model_tiestrength = glm(log(strength.mean)~no.of.papers:Gender+no.of.papers, family=\"gaussian\", data = new) model_tiestrength = glm(log(strength.mean)~no.of.papers:Gender, family=\"gaussian\", data = new) hist(resid(model_tiestrength), main = \"residuals of model with log(tie strength)\") qqnorm(resid(model_tiestrength)) summary(model_tiestrength) summ(model_tiestrength) new$residuals.strength <- resid(model_tiestrength) hist(new$residuals.strength)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get number of papers (and other stats) per gender [CODESPLIT] ddply(new, \"Gender\",summarise, mean = mean(no.of.papers, na.rm = TRUE), median = median(no.of.papers, na.rm = TRUE), sd = sd(no.of.papers, na.rm = TRUE), N = sum(!is.na(no.of.papers)), se = sd / sqrt(N), min = min(no.of.papers, na.rm = TRUE), max = max(no.of.papers, na.rm = TRUE))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create censoring VARIABLE PI status: 1 if PI, 0 if author did not make it to PI [CODESPLIT] new.PI$status.PI = ifelse(!is.na(new.PI$Time.to.PI),1,0)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Gender effect of Gender on time to PI [CODESPLIT] flexgender<-flexsurvreg(Surv(new.PI$time.PI,new.PI$status.PI)~Gender, dist=\"lnorm\", data=new.PI) flexgender", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit_lognormal Lowest AIC, best fit Package flexsurv provides access to additional distributions Also allows plotting options want red lines to match as close as possible to KM curve [CODESPLIT] fit_exp<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"exp\") fit_weibull<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"weibull\") fit_gamma<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"gamma\") fit_gengamma<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"gengamma\") fit_genf<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"genf\") fit_lognormal<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"lnorm\") fit_gompertz<-flexsurvreg(Surv(new$n.years,new$status)~1, dist=\"gompertz\") fit_exp fit_weibull fit_gamma fit_gengamma fit_genf fit_lognormal fit_gompertz plot(fit_exp) plot(fit_weibull) plot(fit_gamma) plot(fit_gengamma) plot(fit_genf) plot(fit_lognormal) plot(fit_gompertz)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generalized gamma distribution is not sig. better fit than lognormal distribution. Justifies using lognormal. AFT model AFT model uses survreg function from Survival package using lognormal distribution according to earlier exploration of the best distribution for the data use survreg for stepwise then flexsurvreg to generate graphs GENDER Examine gender differences first, ignoring any of the social metrics Effect of gender on career length (n.years) [CODESPLIT] flexAFTgender<-flexsurvreg(Surv(n.years,status)~Gender,dist=\"lnorm\",data=new) flexAFTgender plot(flexAFTgender,col=c(\"blue\",\"red\"),ci=T,xlab=\"Years\",ylab=\"Survival probability\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Reverses the factor level ordering for labels after coord_flip() [CODESPLIT] df$labeltext<-factor(df$labeltext, levels=rev(df$labeltext)) df$colour <- c(\"dark grey\",\"dark grey\",\"black\",\"black\",\"dark grey\",\"dark grey\") df_TPICL$labeltext_TPICL <- factor(df_TPICL$labeltext_TPICL, levels=rev(df_TPICL$labeltext_TPICL)) df_TPICL$colour <- c(\"dark grey\",\"dark grey\",\"dark grey\",\"dark grey\",\"dark grey\",\"dark grey\",\"black\",\"black\",\"dark grey\",\"dark grey\",\"dark grey\",\"dark grey\") plotTPICL1<- ggplot(data=df_TPICL, aes(x=labeltext_TPICL, y=estimate_TPICL, ymin=lower_TPICL, ymax=upper_TPICL, col=Gender_TPICL, shape=Gender_TPICL))+ annotate(\"rect\", xmin = 0, xmax = 6.5, ymin = 1, ymax = 5, alpha = .2, fill = \"grey\") + annotate(\"rect\", xmin = 6.5, xmax = 12.7, ymin = 0, ymax = 1, alpha = .2, fill = \"grey\") + geom_point(aes(size=2,col = colour, fill = colour),show.legend=FALSE)+ geom_pointrange(size=0.7, col = df_TPICL$colour, fill = df_TPICL$colour,show.legend=FALSE) + geom_hline(yintercept=1, lty=2) + # Add a dotted line at x=1 after flip geom_vline(xintercept=6.5, lty=1) + # Adds a solid line at y=6.5 after flip to differentiate between top panel = career longevity and bottom panel = time to PI geom_errorbar(aes(ymin=lower_TPICL, ymax=upper_TPICL,width=0.6,cex=1,col=colour, fill = colour),show.legend=FALSE) + scale_x_discrete(breaks=c(1,3,5,7,9,11), labels=c(\"Adj. Network Size\",\"Adj. Tie Strength\",\"Adj. Clustering Coef.\", \"Adj. Network Size\",\"Adj. Tie Strength\",\"Adj. Clustering Coef.\")) + coord_flip() + # Flip coordinates (puts labels on y axis) xlab(\"Label\") + ylab(\"Deceleration factor\") + scale_fill_manual(values=c(\"dark grey\",\"dark grey\",\"black\",\"black\",\"dark grey\",\"dark grey\",\"black\",\"black\",\"dark grey\",\"dark grey\",\"dark grey\",\"dark grey\"))+ scale_colour_manual(values=c(\"black\",\"dark grey\",\"dark grey\",\"dark grey\",\"black\",\"black\")) + scale_shape_manual(values=c(17,16))+ annotate(\"text\",x=7.2, y=5, label=\"Time to become a PI\",size=5,fontface=\"bold\",hjust=1) + annotate(\"text\",x=1, y=5, label=\"Career length\", size=5,fontface=\"bold\",hjust=1) + theme(axis.title.y = element_blank(), # Remove y axis title axis.text.y = element_text(size=12, colour=\"black\"), # Change size of y axis labels axis.text.x = element_text(size=12,colour=\"black\"), # Change size of x axis numbers axis.title.x = element_text(size=14,vjust=0.5,colour=\"black\",face=\"bold\"), # Change size and font of x axis title and move it down a bit panel.grid.major = element_blank(), # Formatting to create blank plot with box around it axis.line = element_line(colour=\"black\"), panel.background = element_rect(colour = \"black\", size=1, fill=NA), legend.title =element_blank())+ guides(colour=FALSE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add label to plot C [CODESPLIT] plotTPICL2<-plotTPICL1+ labs(tag=\"C\")+ theme(plot.tag.position = c(0.19,0.5),plot.tag = element_text(size=14,face=\"bold\"))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Now layout plots and plot them labels for plots A and B added here [CODESPLIT] ggarrange(plotPIstatus,plotTPICL2,heights=c(1,1.7),ncol=1,nrow=2,labels=c(\"A\",\"B\"),label.x=0.18,label.y=0.95)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Display raincloud plot [CODESPLIT] ggplot(dataset_anova, aes(session, score, fill = group)) + geom_rain(alpha = .5, cov = \"group\", rain.side = 'f2x2') + theme(text=element_text(size=20))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a covariance matrix of the means in the placebo group [CODESPLIT] cov_group1 = matrix(c(vcov(cov_fit)[1,1],vcov(cov_fit)[1,3],vcov(cov_fit)[3,1], vcov(cov_fit)[3,3]),2,2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a covariance matrix of the means in the treatment group [CODESPLIT] cov_group2 = matrix(c(vcov(cov_fit)[2,2],vcov(cov_fit)[2,4],vcov(cov_fit)[4,2], vcov(cov_fit)[4,4]),2,2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Contrasting Two vs Fourstakes paradigms Decision generates table of beta coefficients and associated statistics [CODESPLIT] coef.stakes_decision <- data.frame(cbind(summary(m.stakes_decision)$coefficients, m.stakes_decision.CI)) %>% tibble::rownames_to_column('Variable') %>% rename(SE = Std..Error) %>% rename(Z = z.value) %>% rename(P = Pr...z..) %>% rename(CI_L = X2.5..) %>% rename(CI_U = X97.5..) %>% mutate(Variable = factor(Variable, levels = c('fairnessunfair:recodedDecisionreject:stakes4', 'recodedDecisionreject:stakes4', 'fairnessunfair:stakes4', 'fairnessunfair:recodedDecisionreject', 'stakes4', 'recodedDecisionreject', 'fairnessunfair', 'genderfemale', 'centredAge', '(Intercept)' )))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculating the conditional entropy of denomination given designs by authorities separately for each df HIGHER/LOWER Higher denominations [CODESPLIT] authority <- sort(unique(higher$AUTHORITY)) CEDenomination.Designs <- c() HDenominations <- c() NormCEDenomination.Designs <- c() Ncoins <- c() AUTHORITY <- c() for(i in authority){ high_sub <- subset(higher, higher$AUTHORITY == i) denom_high <- as.data.frame(high_sub[,256:309]) motifs_high <- as.data.frame(high_sub[,318:681]) AUTHORITY[i] <- i CEDenomination.Designs[i] <- condentropy(denom_high, motifs_high) HDenominations[i] <- entropy(denom_high) NormCEDenomination.Designs[i] <- condentropy(denom_high, motifs_high) / entropy(denom_high) Ncoins[i] <- nrow(high_sub) } results_higher <- data.frame(\"AUTHORITY\" = AUTHORITY, \"CEDenomination.Designs\" = CEDenomination.Designs, \"Entropy.Denominations\" = HDenominations, \"NormCEDenomination.Designs\" = NormCEDenomination.Designs, \"Ncoins\" = Ncoins) write.csv2(results_higher,\"P3test_byauth_higher.csv\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "destructure string into an array using \"_\" as the delimiter [CODESPLIT] stringArr <- unlist(strsplit(as.character(string), split = \"_\", fixed = T))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "log transform RTs for statistical analysis [CODESPLIT] RTdata$logRT <- log10(RTdata$rt) hist(RTdata$logRT)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit a lienar mixed effects model with all theoretically relevant fixed effects and random intercepts for participant and target context [CODESPLIT] modelRT <- mixed(logRT ~ congruency*soa*mask_contrast + (1| participant) + (1|target_context), data = RTdata, method = \"S\", type=3) modelRT", "target": 1, "target_options": ["no_match", "match"]}
{"input": "pearon correlations within each of these four conditions [CODESPLIT] ca_subset <- CA %>% filter(soa == \"200 ms\" & mc == \"100% contrast\") cor.test(ca_subset$PC, ca_subset$CE) ca_subset <- CA %>% filter(soa == \"400 ms\" & mc == \"100% contrast\") cor.test(ca_subset$PC, ca_subset$CE) ca_subset <- CA %>% filter(soa == \"200 ms\" & mc == \"20% contrast\") cor.test(ca_subset$PC, ca_subset$CE) ca_subset <- CA %>% filter(soa == \"400 ms\" & mc == \"20% contrast\") cor.test(ca_subset$PC, ca_subset$CE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "descriptive statistics of accuracy / errors [CODESPLIT] er_desc <- data_er %>% group_by(participant) %>% summarize(n_correct = sum(acc), n_trials = length(participant))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit a GLMM model with all theoretically relevant fixed effects [CODESPLIT] modelER <- mixed(acc ~ congruency*soa*mask_contrast + (1| participant) + (1|target_context), family = binomial(\"logit\"), data = data_er, method = \"LRT\", type =3) summary(modelER) anova(modelER)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit a GLMM model with all theoretically relevant effects [CODESPLIT] modelER <- mixed(acc ~ soa*mask_contrast + (1| participant) + (1|prime_context), family = binomial(\"logit\"), data = acc_data, method = \"LRT\", type=3) modelER", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate overall proportion [CODESPLIT] prop_overall := N / sum(N)]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "creating variables for lagged values and moving averages [CODESPLIT] delay_rf <- function(v,lag) c(rep(NA,lag),v[1:(length(v)-lag)]) moving_avg <- function(x,n) c(stats::filter(x,rep(1/n,n),sides=1)) res <- lapply(res,transform,cd4Rcd8=cd4/cd8) res <- lapply(res,transform,cd4_ma12=moving_avg(cd4,12),cd8_ma12=moving_avg(cd8,12),cd4Rcd8_ma12=moving_avg(cd4Rcd8,12),cd4_ma24=moving_avg(cd4,24),cd8_ma24=moving_avg(cd8,24),cd4Rcd8_ma24=moving_avg(cd4Rcd8,24), rna_ma12=moving_avg(rna,12),rna_ma24=moving_avg(rna,24)) res <- lapply(res,transform,cd4_lag12=delay_rf(cd4,12),cd4_lag24=delay_rf(cd4,24),cd4_lag36=delay_rf(cd4,36), cd8_lag12=delay_rf(cd8,12),cd8_lag24=delay_rf(cd8,24),cd8_lag36=delay_rf(cd8,36), cd4Rcd8_lag12=delay_rf(cd4Rcd8,12),cd4Rcd8_lag24=delay_rf(cd4Rcd8,24),cd4Rcd8_lag36=delay_rf(cd4Rcd8,36), rna_lag12=delay_rf(rna,12),rna_lag24=delay_rf(rna,24),rna_lag36=delay_rf(rna,36), cd4_ma12_lag12=delay_rf(cd4_ma12,12),cd4_ma24_lag12=delay_rf(cd4_ma24,12),cd4_ma12_lag24=delay_rf(cd4_ma12,24), cd8_ma12_lag12=delay_rf(cd8_ma12,12),cd8_ma24_lag12=delay_rf(cd8_ma24,12),cd8_ma12_lag24=delay_rf(cd8_ma12,24), cd4Rcd8_ma12_lag12=delay_rf(cd4Rcd8_ma12,12),cd4Rcd8_ma24_lag12=delay_rf(cd4Rcd8_ma24,12),cd4Rcd8_ma12_lag24=delay_rf(cd4Rcd8_ma12,24), rna_ma12_lag12=delay_rf(rna_ma12,12),rna_ma24_lag12=delay_rf(rna_ma24,12),rna_ma12_lag24=delay_rf(rna_ma12,24)) res <- lapply(res,transform,haart_lag6=delay_rf(haart,6),haart_lag12=delay_rf(haart,12),hepC_aHCVpos_lag12=delay_rf(hepC_aHCVpos,12),hepC_RNApos_lag12=delay_rf(hepC_RNApos,12), hepC_RNAdet_lag12=delay_rf(hepC_RNAdet,12),hepB_pos_lag12=delay_rf(hepB_pos,12),CDC_lag12_cat=delay_rf(CDC_cat,12),bpn_lag12=delay_rf(bpn,12)) res_df <- rbind.fill(res) res_df$CDC_lag12_cat[res_df$CDC_lag12_cat==1] <- \"A\" res_df$CDC_lag12_cat[res_df$CDC_lag12_cat==2] <- \"B\" res_df$CDC_lag12_cat[res_df$CDC_lag12_cat==3] <- \"C\" res_df$CDC_lag12_cat <- factor(res_df$CDC_lag12_cat,levels=c(\"A\",\"B\",\"C\")) res_df$hepC_aHCVpos_lag12[is.na(res_df$hepC_aHCVpos_lag12)] <- FALSE res_df$hepC_RNApos_lag12[is.na(res_df$hepC_RNApos_lag12)] <- FALSE res_df$hepC_RNAdet_lag12[is.na(res_df$hepC_RNAdet_lag12)] <- FALSE res_df <- transform(res_df,hepC_RNA=(hepC_RNApos | hepC_RNAdet),hepC_RNA_lag12=(hepC_RNApos_lag12 | hepC_RNAdet_lag12)) res_df$bpn_lag12[is.na(res_df$bpn_lag12)] <- FALSE rownames(res_df) <- NULL if(write_df) save(res_df,file=paste(filepath_read_R,\"\\\\monthly_rf_df_\",max_dist,\".RData\",sep='')) print(proc.time()-ptm)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create dummy variables gender [CODESPLIT] Sample2$Dfemale <- recode(Sample2$gender, \"'weiblich'=1;; 'maennlich'=0;; 'divers'=0;; 'keine Angabe'=0\") Sample2$Dmale <- recode(Sample2$gender, \"'maennlich'=1;; 'weiblich'=0;; 'divers'=0;; 'keine Angabe'=0\") Sample2$Ddiverse <- recode(Sample2$gender, \"'divers'=1;; 'weiblich'=0;; 'maennlich'=0;; 'keine Angabe'=0\") Sample2$DNoGenderInd <- recode(Sample2$gender, \"'keine Angabe'=1;;'weiblich'=0;; 'maennlich'=0;; 'divers'=0\") Sample2$DComplianceFully <- recode(Sample2$complianceCurrent, \"'Ja'=1;; 'Ja, teilweise'=0;; 'Nein auch zuvor nicht'=0;; 'Nein, nicht mehr'=0\") Sample2$DCompliancePartly <- recode(Sample2$complianceCurrent, \"'Ja'=0;; 'Ja, teilweise'=1;; 'Nein auch zuvor nicht'=0;; 'Nein, nicht mehr'=0\") Sample2$DComplianceNever <- recode(Sample2$complianceCurrent, \"'Ja'=0;; 'Ja, teilweise'=0;; 'Nein auch zuvor nicht'=1;; 'Nein, nicht mehr'=0\") Sample2$DComplianceNotAnymore <- recode(Sample2$complianceCurrent, \"'Ja'=0;; 'Ja, teilweise'=0;; 'Nein auch zuvor nicht'=0;; 'Nein, nicht mehr'=1\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots Anxiety main effects LISD [CODESPLIT] ggplot(Sample2_Reg,aes(y=Anxiety,x=z_LISD_State_F1))+ geom_point(color = \"indianred4\")+geom_smooth(method=\"lm\", color = \"black\", size = 0.5)+theme_classic() ggplot(Sample2_Reg,aes(y=Anxiety,x=z_LISD_Trait_F1))+ geom_point(color = \"indianred4\")+geom_smooth(method=\"lm\", color = \"black\", size = 0.5)+theme_classic() ggplot(Sample2_Reg,aes(y=Anxiety,x=z_LISD_Trait_F2))+ geom_point(color = \"indianred4\")+geom_smooth(method=\"lm\", color = \"black\", size = 0.5)+theme_classic()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots Depressed plot preparation: divide z_age into SD, mean, +SD [CODESPLIT] attach(Sample2_Reg) Sample2_Reg$age_3groups <- case_when(z_age > mean(z_age)+sd(z_age) ~ \"high\", z_age < mean(z_age)+sd(z_age) & z_age > mean(z_age)-sd(z_age) ~ \"mean\", z_age < mean(z_age)-sd(z_age) ~ \"low\") detach(Sample2_Reg)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plots Manuscript Plots Manuscript with raw WLISD scores Anxiety [CODESPLIT] plot1_1 <- Sample2_Reg %>% ggplot(aes(x = LISD_State_F1, y = Anxiety)) + geom_point(color = \"#00AFBB\")+ geom_smooth(method = lm, color = \"black\", size = 0.5)+ xlim(1,5)+ ylim(-2,3)+ labs(x = \"lonely & isolated (State 1)\", y = \"Anxiety (z)\", Title = \"State Factor 1\")+ theme_classic() plot2_1 <- Sample2_Reg %>% ggplot(aes(x = LISD_State_F2, y = Anxiety)) + geom_point(color = \"#55C667FF\")+ geom_smooth(method = lm, color = \"black\", size = 0.5)+ xlim(1,5)+ ylim(-2,3)+ labs(x = \"supported & connected (State 2)\", y = \"Anxiety (z)\", Title = \"State Factor 2\")+ theme_classic() plot3_1 <- Sample2_Reg %>% ggplot(aes(x = LISD_Trait_F1, y = Anxiety)) + geom_point(color = \"#E7B800\")+ geom_smooth(method = lm, color = \"black\", size = 0.5)+ xlim(1,5)+ ylim(-2,3)+ labs(x = \"loneliness & isolation (Trait 1)\", y = \"Anxiety (z)\", Title = \"Trait Factor 1\")+ theme_classic() plot4_1 <- Sample2_Reg %>% ggplot(aes(x = LISD_Trait_F2, y = Anxiety)) + geom_point(color = \"#FC4E07\")+ geom_smooth(method = lm, color = \"black\", size = 0.5)+ xlim(1,5)+ ylim(-2,3)+ labs(x = \"sociability & sense of belonging (Trait 2)\", y = \"Anxiety (z)\", Title = \"Trait Factor 1\")+ theme_classic() plot1_1 plot2_1 plot3_1 plot4_1", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a line graph articles per year [CODESPLIT] ggplot(ArticlesbyYear, aes(x=Year, y=n, group = 1)) + geom_line(color=\"orange\") + labs(y = \"Number of Articles\", angle = 45) + theme_minimal(base_size = 12)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a barplot articles per year [CODESPLIT] ggplot(ArticlesbyYear, aes(x = Year, y = n)) + geom_bar(stat = \"identity\", color = \"steelblue3\", fill = \"steelblue3\") + theme_minimal(base_size = 12) + labs(y = \"Articles\", angle = 45) + geom_text(aes(label = n), vjust =\"center\", size=3, hjust = \"center\", nudge_y = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a barplot articles per journal [CODESPLIT] ggplot(Journals, aes(x = reorder(Journal, Percent), y = Percent)) + geom_bar(stat = \"identity\", color = \"steelblue3\", fill = \"steelblue3\") + coord_flip() + theme_minimal(base_size = 11) + labs(y = \"Percent\", x = \"Journal\") + geom_text(aes(label = Percent), vjust =\"center\", size=3, hjust = \"center\", nudge_y = 0.01)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create barplot Subfields Percents counts [CODESPLIT] ggplot(subfield_count, aes(x = reorder(Subfield, Percent) , y = Percent, fill = Subfield)) + geom_bar(stat = \"identity\") + scale_fill_brewer(palette = \"Blues\", guide = FALSE) + coord_flip() + theme_minimal(base_size = 13) + labs(y = \"Percent of Articles\", x = \"Subfield\") + geom_text(aes(label = Percent), vjust =\"center\", size=3, hjust = \"center\", nudge_y =.1) + scale_y_continuous(labels = scales::percent, limits = c(0,1))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "authorlevel count and proportion for all gender identity categories [CODESPLIT] gender_author <- author_data %>% subset(!is.na(Gender.apsa)) %>% dplyr::summarize(count = c(sum(male_apsa, na.rm=T), sum(female_apsa, na.rm=T), sum(nonbinary_apsa, na.rm=T))) gender_author <- gender_author %>% mutate(proportion = round(count / sum(count), 2)) %>% mutate(gender = c(\"male\", \"female\", \"nonbinary\")) %>% dplyr::select(gender, count, proportion)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate count and proportion of gender author structure [CODESPLIT] gender_article_dedup <- gender_article %>% distinct(article_title, single_authored_male, single_authored_female, co_authored_male, co_authored_female, co_authored_mixed) gender_article_count <- data.frame(matrix(NA, nrow = 5, ncol = 3)) colnames(gender_article_count) <- c(\"Author_Gender\", \"Frequency\", \"Percent\") gender_article_count$Author_Gender <- c(\"Single Authored Male\", \"Single Authored Female\", \"Co-authored Male\", \"Co-authored Female\", \"Co-authored Mixed Gender\") gender_article_count$Frequency <- c(sum(gender_article_dedup$single_authored_male, na.rm=T), sum(gender_article_dedup$single_authored_female, na.rm=T), sum(gender_article_dedup$co_authored_male, na.rm=T), sum(gender_article_dedup$co_authored_female, na.rm=T), sum(gender_article_dedup$co_authored_mixed, na.rm=T)) gender_article_count$Percent <- formattable::percent(gender_article_count$Frequency/sum(gender_article_count$Frequency), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "articlelevel authorship structure for race / ethnic identity categories [CODESPLIT] race_ethnicity_article <- race_ethnicity_article %>% group_by(article_title) %>% mutate(white_authors = ifelse(mean(white) == 1, 1, 0), black_authors = ifelse(mean(black) == 1, 1, 0), east_asian_authors = ifelse(mean(east_asian) == 1, 1, 0), south_asian_authors = ifelse(mean(south_asian) == 1, 1, 0), latino_authors = ifelse(mean(latino) == 1, 1, 0), mena_authors = ifelse(mean(mena) == 1, 1, 0), native_authors = ifelse(mean(native) == 1, 1, 0), pacific_authors = ifelse(mean(pacific) == 1, 1, 0), other_authors = ifelse(mean(other) == 1, 1, 0), mixed_authors = ifelse(mean(white) < 1 & n() > 1, 1, 0))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "count of articles that generated data using experimental techniques (includes articles that use both, percentage calculated using total empirical articles) [CODESPLIT] GenerateData[2,2] <- sum(MMCPSR_emp$EHPdata) GenerateData[2,3] <- sum(MMCPSR_emp$EHPdata)/nrow(MMCPSR_emp)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number (%) of articles drawing on data collected via Survey (alone + in combo. w/other techniques) Raw Count/percent alone [CODESPLIT] SurveySoloCount <- sum(MMCPSR_emp$`Ethnography / participant observation` ==0 & MMCPSR_emp$`Interviews/focus groups` == 0 & MMCPSR_emp$Survey == 1 & MMCPSR_emp$EHPdata == 0 & MMCPSR_emp$gendataNHP == 0 & MMCPSR_emp$`Employed data/information from pre-existing primary or secondary sources`==0) SurveySoloCount formattable::percent(SurveySoloCount/nrow(MMCPSR_emp), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number (%) of articles using Field experiments (alone + in combo. w/other techniques) Count/percent alone [CODESPLIT] FieldExpSoloCount <- sum(MMCPSR_emp$`Survey experiment` ==0 & MMCPSR_emp$Field == 1 & MMCPSR_emp$Lab == 0 & MMCPSR_emp$OHPdata == 0 & MMCPSR_emp$gendataNHP == 0 & MMCPSR_emp$`Employed data/information from pre-existing primary or secondary sources`==0) FieldExpSoloCount formattable::percent(FieldExpSoloCount/nrow(MMCPSR_emp), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number (%) of articles using data generated through interaction with Dom gov [CODESPLIT] sum(MMCPSR_emp$`EHP - Domestic government` == 1 | MMCPSR_emp$`OHP - Domestic government`==1) formattable::percent(sum(MMCPSR_emp$`EHP - Domestic government` == 1 | MMCPSR_emp$`OHP - Domestic government`==1) /nrow(MMCPSR_emp), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number (%) of articles using data generated through interaction with Media [CODESPLIT] sum(MMCPSR_emp$`EHP - Media` == 1 | MMCPSR_emp$`OHP - Media`==1) formattable::percent(sum(MMCPSR_emp$`EHP - Media` == 1 | MMCPSR_emp$`OHP - Media`==1)/nrow(MMCPSR_emp), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "probit and marginal effects experimental data vs time [CODESPLIT] out.1.probit <-glm(EHPdata~Year, data = MMCPSR_emp, family = binomial(link = \"probit\")) summary(out.1.probit) probitmfx(out.1.probit, data = MMCPSR_emp)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create correlation matrix author gender vs methods categories [CODESPLIT] MethodGender_cor <- rcorr(as.matrix(MethodGender_sub))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "DV Formal modeling only OLS Modeling only vs time [CODESPLIT] out.12 <- lm(MMCPSR_emp$ModelingOnly ~ MMCPSR_emp$Year) summary(out.12) out.15$coefficients[2]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Conclusion Policy Recommendations count and percent of all articles [CODESPLIT] sum(MMCPSR_Data$`Policy Recommendation`) formattable::percent(sum(MMCPSR_Data$`Policy Recommendation`)/nrow(MMCPSR_Data), digits = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "DV time OLS policy recommendations vs time [CODESPLIT] out.21 <- lm(MMCPSR_emp$`Policy Recommendation`~MMCPSR_emp$Year) summary(out.21) out.21$coefficients[2] out.21b <- lm(MMCPSR_emp$`Policy Recommendation`~as.factor(MMCPSR_emp$Year)) summary(out.21b) plot_coefs(out.21b) plot_coefs(lm(MMCPSR_emp$`Policy Recommendation`~ 0 + as.factor(MMCPSR_emp$Year)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "scaling all the variables of interest (between 01) for a composite score [CODESPLIT] data8$gincdif_s = rescale(data8$gincdif) data8$smdfslv_s = rescale(data8$smdfslv) data8$sbstrec_r = 6 - data8$sbstrec # reverse scores first data8$sbstrec_r_s = rescale(data8$sbstrec_r)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create one index of political engagement using: interest, media reception, and political action [CODESPLIT] polit_eng = c(\"polintr_r_s\", \"nwspol_s\", \"polit_action_s\") data8$polit_eng = rowMeans(data8[polit_eng], na.rm = TRUE) mean(data8$polit_eng, na.rm = TRUE) # 0.22 sd(data8$polit_eng, na.rm = TRUE) # 0.14 psych::alpha(data8[polit_eng]) # 0.39 data8$polit_eng_c = c(scale(data8$polit_eng, center = TRUE, scale = FALSE)) # center", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate values scores for PVQ recoding PVQ items (ESS uses 1 as \"like me\" and 6 \"totally not like me) [CODESPLIT] data8$ipcrtiv_r = 7 - data8$ipcrtiv data8$imprich_r = 7 - data8$imprich data8$ipeqopt_r = 7 - data8$ipeqopt data8$ipshabt_r = 7 - data8$ipshabt data8$impsafe_r = 7 - data8$impsafe data8$impdiff_r = 7 - data8$impdiff data8$ipfrule_r = 7 - data8$ipfrule data8$ipudrst_r = 7 - data8$ipudrst data8$ipmodst_r = 7 - data8$ipmodst data8$ipgdtim_r = 7 - data8$ipgdtim data8$impfree_r = 7 - data8$impfree data8$iphlppl_r = 7 - data8$iphlppl data8$ipsuces_r = 7 - data8$ipsuces data8$ipstrgv_r = 7 - data8$ipstrgv data8$ipadvnt_r = 7 - data8$ipadvnt data8$ipbhprp_r = 7 - data8$ipbhprp data8$iprspot_r = 7 - data8$iprspot data8$iplylfr_r = 7 - data8$iplylfr data8$impenv_r = 7 - data8$impenv data8$imptrad_r = 7 - data8$imptrad data8$impfun_r = 7 - data8$impfun", "target": 1, "target_options": ["no_match", "match"]}
{"input": "evaluate goodness of fit of psychometric function using deviance test [CODESPLIT] d_fit <-{} for(i in unique(d$Subject)){ m_1 <- glm(Response ~ JumpSize * Velocity +BlinkDuration , d[d$Subject==i & d$cond1==1,], family=binomial(probit)) m_0<- glm(Response ~ Velocity +BlinkDuration , d[d$Subject==i & d$cond1==1,], family=binomial(probit)) LRT <- anova(m_0, m_1,test=\"LRT\") d_fit <- rbind(d_fit, data.frame(id=i, D=LRT$Deviance[2], df=LRT$Df[2], p=LRT$`Pr(>Chi)`[2])) } print(d_fit,digits=2,row.names=F) round(d_fit$p,digits=10)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "use lmList to fit individual GLM models in one go [CODESPLIT] mo <- lmList(Response ~ cond1 * JumpSize * Velocity +bdur:cond1 | Subject, d, family=binomial(probit))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "visualize Bayes factors and parameter estimates [CODESPLIT] par(mfrow=c(1,2)) slope_bdur <- B[,8] / (((B[,3]+B[,5]) + (B[,3]+B[,7]+B[,5]+B[,9]))/2) slope_bayes <- outB$beta_5 / (((outB$beta_2) + (outB$beta_2+outB$beta_4))/2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "sanity check to ensure that we get same estimates from frequentist or Bayesian analysis [CODESPLIT] plot(slope_bdur, slope_bayes, ylab=\"slope (Bayesian)\",xlab=\"slope (frequentist)\", pch=19) abline(a=0,b=1,lty=2) cor.test(slope_bdur, slope_bayes)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "n: sample size of each study simtot: number of simulations seed 1 Compute d33, ncp33 and df for the original sample [CODESPLIT] d33=pwr.t.test(n=n,sig.level=0.05,power=1/3,type=\"two.sample\")$d ncp33=sqrt(n/2)*d33 df=2*n-2", "target": 1, "target_options": ["no_match", "match"]}
{"input": "point estimate p hat (just mean average of coefficients) [CODESPLIT] pe <-mean(cors)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "next, we use the bridge_sampler() function which will stabilize the calculated BFs see Schad et al 2022 https://psycnet.apa.org/doi/10.1037/met0000472 [CODESPLIT] model.intox.seq.bf <- bridge_sampler(model.intox.seq, silent = TRUE) model.intox.seq.null.bf <- bridge_sampler(model.intox.seq.null, silent = TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "figure out which columns contain NAs [CODESPLIT] missing_cols <- colSums(is.na(data.conseq)) > 0 print(names(data.conseq)[missing_cols])", "target": 1, "target_options": ["no_match", "match"]}
{"input": "age recored as 1 18, to correct we added 17 to all age variables [CODESPLIT] data$age <- data$age + 17", "target": 1, "target_options": ["no_match", "match"]}
{"input": "User language variable into a factor [CODESPLIT] data$UserLanguage <- ifelse(data$UserLanguage == \"EN\", \"EN\", \"NL\") data$UserLanguage <- as.factor(data$UserLanguage) contrasts(data$UserLanguage) <- c(0, 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Test assumptions MANOVA Test whether residuals are normally distributed [CODESPLIT] df$pc1.residuals = lm(pc1~condition.breach, data=df)$residuals df$pc2.residuals = lm(pc2~condition.breach, data=df)$residuals df$pc3.residuals = lm(pc3~condition.breach, data=df)$residuals df$pc4.residuals = lm(pc4~condition.breach, data=df)$residuals shapiro.test(df$pc1.residuals) shapiro.test(df$pc2.residuals) shapiro.test(df$pc3.residuals) shapiro.test(df$pc4.residuals)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot effect of first and third principal component [CODESPLIT] eigenfunctions.df = read.table(file=\"../data/final/eigenfunctions_breach.dat\", header=T) fig.df1 <- data.frame(HR = c(eigenfunctions.df$pc1*-0.0305, eigenfunctions.df$pc1*0.1355), Condition = rep(c(rep(\"Breach\", 51), rep(\"Fulfilment\", 51))), Time=rep(seq(0,1, by=1/50),2), pc=\"Principal component 1\") fig.df2 <- data.frame(HR = c(eigenfunctions.df$pc3*0.0278, eigenfunctions.df$pc3*-0.0617), Condition = rep(c(rep(\"Breach\", 51), rep(\"Fulfilment\", 51))), Time=rep(seq(0,1, by=1/50),2), pc=\"Principal component 3\") fig.df = rbind(fig.df1, fig.df2) fig <- ggplot(fig.df, aes(x=Time, y =HR, linetype=Condition))+ geom_line()+ geom_vline(xintercept=.20, linetype=\"dashed\")+ facet_grid(~pc)+ theme_bw() fig ggsave(file=\"../figures/figure3.png\", width=8, height=6)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function pcor2beta gives you data from a partial correlation/network input pcor a partial correlation matrix / network output a matrix of betas, each column corresponds to a dependent variable so that you can get predicted values by a matrix multiplication in the form betas %*% data [CODESPLIT] pcor2beta <- function(pcor) { require(psych) require(corpcor) diag(pcor) <- 1 p <- ncol(pcor) betas <- matrix(0, ncol = p, nrow = p) for(i in 1:p) betas[-i,i] <- matReg(y = i, x = seq(p)[-i], C = pcor2cor(pcor))$beta betas[abs(betas) < 1e-13] <- 0 betas }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "function R2 gives you two different types of R2 and of predicted values input pcor a partial correlation matrix / network output a matrix of betas, each column corresponds to a dependent variable so that you can get predicted values by a matrix multiplication in the form betas %*% data this function gives you R2 and predicted values from betas + data two types of R2 and predicted values are considered: R2_orig and predicted_orig use beta weights directly implied by the graphical lasso regularization R2_refit and predicted_refit use only the sparsity pattern of the network and then refit linear regression using the network only for prediction but not for shrinkage (this one should do better in terms of prediction) refit: logical, regulates whether R2_refit and predicted_refit are computed. set it to FALSE to speed up computations [CODESPLIT] R2 <- function(betas, dt, refit = TRUE) { dt <- data.frame(scale(dt)) out <- list() p <- ncol(betas)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "refit the model considering the sparsity pattern inicated by the network first fit regression using the sparsity indciated in the matrix of betas [CODESPLIT] betas_refit <- matrix(0, ncol = p, nrow = p) for(i in 1:p) { if(any(betas[,i] != 0)) { fit <- lm(dt[,i] ~ as.matrix(dt[,betas[,i] != 0])) betas_refit[betas[,i] != 0, i] <- fit$coefficients[-1] } else betas_refit[,i] <- 0 } predicted <- as.matrix(dt) %*% betas_refit", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit the data using the lm function [CODESPLIT] bcafit <- lm(Abs562 ~ BSA, data = bca)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "rename the columns for display [CODESPLIT] colnames(samptab) <- c(\"Absorbance\", \"Concentration (mg/mL)\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "departure time (12 hours so we can calculate easy the mean) [CODESPLIT] d1[, time_ := strftime(datetime_ - 60*60*12, format=\"%H:%M:%S\")] d1[, time_ := as.POSIXct(time_, format=\"%H:%M:%S\")]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "This creates a variable \"period\" in the dataframe myDataISO: [CODESPLIT] myDataISO$period<-gl(2, 1202, labels = c(\"first\", \"second\")) # We created 2 sets of 1202 scores, the labels option then specifies the names to attach to these 2 sets, which correspond to the levels of \"period\" (first measurement and follow-up).", "target": 1, "target_options": ["no_match", "match"]}
{"input": "MIXED DESIGNS AS A GLM Setting contrasts for quarantine subperiod: [CODESPLIT] myDataISO$quar.duration<-as.factor(myDataISO$quar.duration) is.factor(myDataISO$quar.duration) basvs10days<-c(0,1,0,0) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 10-days duration basvs50days<-c(0,0,1,0) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 50-days duration basvs103days<-c(0,0,0,1) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 103-days duration contrasts(myDataISO$quar.duration)<-cbind(basvs10days, basvs50days, basvs103days) myDataISO$quar.duration # To check we setted the contrasts correctly myDataISOIMP2$quar.duration<-as.factor(myDataISOIMP2$quar.duration) is.factor(myDataISOIMP2$quar.duration) basvs10days<-c(0,1,0,0) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 10-days duration basvs50days<-c(0,0,1,0) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 50-days duration basvs103days<-c(0,0,0,1) # this compares the baseline (prior to quarantine) to a quarantine sub.period of up to 103-days duration contrasts(myDataISOIMP2$quar.duration)<-cbind(basvs10days, basvs50days, basvs103days) myDataISOIMP2$quar.duration # To check we setted the contrasts correctly", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculating effect sizes library(DSUR.noof) We got effect sizes of meaningful predictors by executing: rcontrast(t,df) periodsecond [CODESPLIT] rcontrast(-2.685100, 97) rcontrast(-5.280908, 1201)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Simulate time varying variable X [CODESPLIT] data.X = expand.grid(Obs=1:T.obs,ID=1:N.dyad) var.diag.X = c(sigma.XF,sigma.XM) Sigma.X = diag(length(var.diag.X)) Sigma.X[lower.tri(Sigma.X, diag=FALSE)] = rho.X Sigma.X = pmax(Sigma.X, t(Sigma.X), na.rm=TRUE) Sigma.X = diag(var.diag.X)%*%Sigma.X%*%diag(var.diag.X) data.X = cbind(data.X,mvrnorm(N.dyad*T.obs,c(mu.XF,mu.XM),Sigma.X)) colnames(data.X) = c('Obs','ID','X.F','X.M') X.F = data.X[,'X.F'] X.M = data.X[,'X.M'] D.dyad = rbinom(N.dyad*T.obs,1,prob.D) data.X = cbind(data.X,D.dyad)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "left_join remaining_uncertain_matched to author_data, subset to remove non [CODESPLIT] author_data <- left_join(author_data, remaining_uncertain_matched, by = c(\"index\" = \"index\")) author_data$match <- as.integer(author_data$match)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remerge (left_join) author_names with author_data [CODESPLIT] author_data <- left_join(test, author_data, by = c(\"Contact.apsa\", \"NameSort.apsa\")) %>% subset(duplicated(Contact.apsa) == FALSE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number of subjects in each group [CODESPLIT] n.group = c(rep(0,N.0),rep(1,N.1)) n.group = c(rep(0,N.0),rep(1,N.1)) n.group = c(rep(0,N.0),rep(1,N.1)) n.group = c(rep(0,N.0),rep(1,N.1))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create variables days, beeps per day and Z [CODESPLIT] data.IL = expand.grid(Time=1:T,Z=n.group) data.IL = expand.grid(Time=1:T,W=W.i) data.IL = expand.grid(Time=1:T,subjno=1:N) data.IL = expand.grid(Time=1:T,subjno=1:N) data.IL = expand.grid(Time=1:T,Z=n.group) data.IL = expand.grid(Time=1:T,Z=n.group) data.IL = expand.grid(Time=1:T,W=W.i) data.IL = expand.grid(Time=1:T,W=W.i) data.IL = expand.grid(Time=1:T,subjno=1:N) data.IL = expand.grid(Time=1:T,Z=n.group) data.IL = expand.grid(Time=1:T,W=W.i)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create variable subjno [CODESPLIT] subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) Z = data.IL$Z subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) W = data.IL$W subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) Z = data.IL$Z subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) Z = data.IL$Z subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) W = data.IL$W subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) W = data.IL$W subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) Z = data.IL$Z subjno = expand.grid(1:T,1:N)[,2] data.IL = cbind(subjno,data.IL) W = data.IL$W", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create lag Y variable within days for each individual [CODESPLIT] Ylag = lag.Y(data) data = data.frame(cbind(data,Ylag)) } Ylag = lag.Y(data) data = data.frame(cbind(data,Ylag)) } Ylag = lag.Y(data) data = data.frame(cbind(data,Ylag)) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Parameters of random intercept and random slope are generated from a Beta distribution Parameters of Group0 Stationarity condition: sigma.v1 < sqrt(1b10^2) [CODESPLIT] if (sigma.v1 > sqrt(1-b10^2)) {stop('To ensure that the model in Group 0 is stationary check that standard deviation of the random slope is smaller than sqrt(1-b10^2) where b10 is the fixed autorregressive effect')} mu.beta.0 = (b10+1)/2 sigma.beta.0 = sigma.v1/2 alpha.beta.0 = mu.beta.0^2*(((1-mu.beta.0)/sigma.beta.0^2) - (1/mu.beta.0)) beta.beta.0 = alpha.beta.0*((1/mu.beta.0) - 1) gamma.01 = rbeta(N.0, alpha.beta.0, beta.beta.0)*2-1 gamma.00 = sigma.v0*(rho.v*scale(gamma.01)+sqrt(1-rho.v^2)*rnorm(N.0)) + rep(b00,N.0)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Parameters of random intercept and random slope are generated from a Beta distribution [CODESPLIT] mu.beta.1 = (b10+b11.W*W.i[i]+1)/2 sigma.beta.1 = sigma.v1/2 alpha.beta.1 = mu.beta.1^2*(((1-mu.beta.1)/sigma.beta.1^2) - (1/mu.beta.1)) beta.beta.1 = alpha.beta.1*((1/mu.beta.1) - 1) gamma.1 = rbeta(1, alpha.beta.1, beta.beta.1)*2-1 gamma.0 = sigma.v0*((rho.v*(gamma.1-(b10+b11.W*W.i[i]))/sigma.v1) + sqrt(1-rho.v^2)*rnorm(1)) + b00+b01.W*W.i[i] if (Ylag.center == TRUE){ AR.epsilon = list(order=c(1,0,0), ar=gamma.1, include.mean = FALSE) Y[which(data.IL$subjno==i)] = arima.sim(n=T,AR.epsilon)*sqrt(1-gamma.1^2)*sigma + gamma.0 } if (Ylag.center == FALSE){ AR.epsilon = list(order=c(1,0,0), ar=gamma.1, include.mean = FALSE) Y[which(data.IL$subjno==i)] = arima.sim(n=T,AR.epsilon)*sqrt(1-gamma.1^2)*sigma + gamma.0/(1-gamma.1) }}", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Get statistics for observed networks [CODESPLIT] obs_df <- GetNetStats(group_fit$networks, group_fit$formula, \"model\") colnames(obs_df) <- stat_labels obs_df <- obs_df %>% mutate(n=1:200, Group=rep(c(\"Young\",\"Old\"), each=100)) %>% melt(measure.vars=stat_labels, variable.name=\"Stat\") obs_df <- rbind(obs_df, data.frame(n=1:200, Group=rep(c(\"Young\",\"Old\"), each=100), Stat = \"Local efficiency\", value = local_eff), data.frame(n=1:200, Group=rep(c(\"Young\",\"Old\"), each=100), Stat = \"Global efficiency\", value = global_eff)) sim_df <- obs_df[0, ] young_nets <- group_fit$networks[1:100] for (i in 1:n_sample){ y <- young_nets[[sample(length(young_nets),1)]] myformula <- statnet.common::nonsimp_update.formula(group_fit$formula, y ~., from.new = \"y\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Source functions that allow to easily draw nice visualizations [CODESPLIT] source(\"visual_functions_all.R\") source(\"visualize_niceplot.R\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Test H1d. \"Compliance with behavioural guidelines\" by Countries random intercept (country) + the fixed effect of country for all coefficients, priors ~ Cauchy (0, 1) [CODESPLIT] prior.coef <- brms::prior(cauchy(0,1),class='b')", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate total number of language selected by each subj. We will use this to compute a mean weight [CODESPLIT] sum.home.languages <- adj.home %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, language) %>% dplyr::summarise(n=any(val==1)) %>% dplyr::summarise(n_languages=sum(n)) sum.fam.languages <- adj.fam %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, language) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_languages=sum(n)) sum.social.languages <- adj.social %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, language) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_languages=sum(n)) sum.work.languages <- adj.work %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, language) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_languages=sum(n))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Sum up the adjacency values for each subject for each topictopic pairing, merge with the number of contexts containing responses, and compute mean [CODESPLIT] sum.adj.home <- adj.home %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.home.languages) %>% dplyr::mutate(avg=sum/n_languages) %>% ungroup() sum.adj.fam <- adj.fam %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.fam.languages) %>% dplyr::mutate(avg=sum/n_languages) %>% ungroup() sum.adj.school <- adj.school %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.school.languages) %>% dplyr::mutate(avg=sum/n_languages) %>% ungroup() sum.adj.social <- adj.social %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.social.languages) %>% dplyr::mutate(avg=sum/n_languages) %>% ungroup() sum.adj.work <- adj.work %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.work.languages) %>% dplyr::mutate(avg=sum/n_languages) %>% ungroup() sum.adj.dom <- adj.dom %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.dom.context) %>% dplyr::mutate(avg=sum/n_context) %>% ungroup() sum.adj.non <- adj.non %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, topic,topic2) %>% dplyr::summarise(sum=sum(val)) %>% dplyr::left_join(sum.non.context) %>% dplyr::mutate(avg=sum/n_context) %>% ungroup()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate total number of languages selected by each subj. We will use this to compute a mean weight [CODESPLIT] sum.school.languages <- adj.school %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, language) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_languages=sum(n))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "NETWORK SIZE (i.e., how many topics are used in each network?) [CODESPLIT] c.size.long <- contexts.mean %>% select(subject, contains(\"networkSize\")) %>% gather(context, network.size, contains(\"networkSize\")) c.size.long$context = factor(c.size.long$context, levels = c(\"work.networkSize\", \"school.networkSize\", \"home.networkSize\", \"fam.networkSize\", \"social.networkSize\"), labels = c(\"Work\", \"School\", \"Home\", \"Family\", \"Social\")) c.size.long = c.size.long %>% filter(!is.na(network.size)) c.size.summary = convenience::sem(c.size.long, dv = network.size, id = subject, context)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate total number of contexts selected by each subj. We will use this to compute a mean weight [CODESPLIT] sum.dom.context <- adj.dom %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, context) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_context=sum(n)) sum.non.context <- adj.non %>% gather(topic2, val, chitchat:gossip) %>% dplyr::group_by(subject, context) %>% dplyr::summarise(n = any(val == 1)) %>% dplyr::summarise(n_context=sum(n))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Language network stats NETWORK WEIGHT (i.e., how many contexts is each topictopic pair used in this language?) [CODESPLIT] language.wide <- sum.adj.dom %>% dplyr::full_join(sum.adj.non, by = c(\"subject\", \"topic\", \"topic2\")) %>% select(-contains(\"n_context\")) %>% select(-contains(\"avg\")) names(language.wide) <- c(\"subject\",\"topic\", \"topic2\", \"Dominant Language\", \"Non-dominant Language\") language.long <- gather(language.wide, language, weight, \"Dominant Language\":\"Non-dominant Language\") language.long = language.long %>% filter(!is.na(weight)) l.weight.summary = convenience::sem(language.long, dv = weight, id = subject, language) language.aov <- aov(weight ~ language, data = language.long) summary.aov(language.aov) # sig***", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Mutate targets to uppercase [CODESPLIT] SPD_all <- SPD_all %>% mutate(target = toupper(target)) dat <- dat %>% mutate(target = toupper(target))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute Zscores separately for each participant and each session [CODESPLIT] group_by(Subject, Session) %>% mutate(Ztarget.RT = scale(target.RT)) %>% ungroup() %>%", "target": 1, "target_options": ["no_match", "match"]}
{"input": "look up pchange in table based on reward (yes/no) and confidence unselected advisor [CODESPLIT] sel.row <- subset(p_switch, Reward == reward & Conf.unsel.adv == unselect.adv.conf) sel.row pswitch <- sel.row$Changed pswitch", "target": 1, "target_options": ["no_match", "match"]}
{"input": "exploratory bifactor analysis [CODESPLIT] fa(resp, fm=\"ml\", cor=\"tet\", nfactors = 5, rotate = \"bifactor\", correct = 0)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "estimating Rasch Tree models for gender & age [CODESPLIT] gender.dif<-raschtree(dat~gender, data=bdat, deriv=\"numeric\", alpha=.01, bonferroni=TRUE) age.dif<-raschtree(dat~age, data=bdat, deriv=\"numeric\", minsize=400, alpha=.01, bonferroni=TRUE) summary(gender.dif) summary(age.dif) plot(gender.dif) plot(age.dif)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "set those study grades to NA which are outside the range of the grading system [CODESPLIT] mSI$crt.gru_s_w2[mSI$crt.gru_s_w2 == 0] <- NA mSI$crt.gru_s_w2[mSI$crt.gru_s_w2 > 6] <- NA", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compute and save descriptives statistics of variables before aggregation and standardization [CODESPLIT] descriptives <- round(select(psych::describe(mSI_descr), n, min, max, mean, sd),2) write.table(descriptives, file=\"Descriptives/descriptives_Sample_E_mSI.dat\", sep=\"\\t\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "mean ratings of targets and decoys by condition [CODESPLIT] tapply(dat.long$rating, list(dat.long$condition, dat.long$target), mean) tapply(dat.long$rating, list(dat.long$condition, dat.long$target), function(x) sd(x)/sqrt(length(x)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "paired contrasts on estimated marginal means to unpack interaction [CODESPLIT] emm_options(lmer.df = \"satterthwaite\") fitlmer2.em <- emmeans::emmeans(fitlmer2, specs = ~ target*condition | target) fitlmer2.em # estimated marginal means contrast(fitlmer2.em, method = \"trt.vs.ctrl\", adjust = \"none\") # paired contrasts with no correction confint(contrast(fitlmer2.em, method = \"trt.vs.ctrl\", adjust = \"none\")) # 95% confidence intervals contrast(fitlmer2.em, method = \"trt.vs.ctrl\", adjust = \"holm\") # paired contrasts with holm correction s <- as.data.frame(summary(fitlmer2.em))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Statistical tests: ECT Choice score vs Training criteria Did mice reduce their choice score in the test compared to 80% criteria? using JBTxECT_id.csv dataset [CODESPLIT] wilcox.test(md_id$choice_score[md_id$rat_bedding != \"mixed\"], mu = 0.6) # 80% big = NCT score 0.6", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Make figures and tables check for correlaitons between individuals [CODESPLIT] chart.Correlation(ind_res[[14]][,1:6] , histogram=TRUE, pch=19)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "define plot styling pp_color for predictive draw lines, set in plotting_style.R [CODESPLIT] cat(\"\\n\\nPlotting posterior predictive checks...\\n\") cat(\"Ignore coordinate system and colour warnings;;\", \"these are expected behavior\\n\") chains <- rstan::extract(fit) n_checks = length(chains$titer_rep_censored[, 1]) sample_checks = sample(1:n_checks, n_to_plot) print(sample_checks) print(names(chains)) cat(sprintf(\"\\nPlotting pp checks...\\n\\n\")) real_titers <- dat$log10_titer rep_titers <- chains$titer_rep_censored[sample_checks, ] plot_upper <- pp_check(10^(real_titers), yrep = 10^(rep_titers), ## convert to RML titer fun = ppc_dens_overlay, alpha = 0.2) + ggtitle(\"Predictive checks\") + scale_color_manual(name=\"\", labels = c(\"real data\", \"posterior predictive draws\"), values = c(\"black\", pp_color)) + scale_x_continuous( trans = 'log10', labels = trans_format('log10', math_format(10^.x))) + coord_cartesian(xlim = c(10^0.5, 10^6.5)) + expand_limits(x = 0.5, y = 0) + theme_project(base_size = 30)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "extract relevant data & calculate Brier score for each participant [CODESPLIT] lay.data <- NULL for (i in 1:nsubjects) { study.order <- as.numeric(unlist(strsplit(prediction.data$preorder[i], split=\"|\", fixed = TRUE)))[2:28] label <- ifelse(prediction.data$Conditie[i]==1,\"des.\",\"bf.\") subject.data <- NULL for (j in 1:nstudies){ understanding <- eval(parse(text = paste0(\"prediction.data$`\",study.order[j],\"_\",label,\"0`\" )))[i] # 1 = did not understand, otherwise: NA replication.belief <- eval(parse(text = paste0(\"prediction.data$`\",study.order[j],\"_\",label,\"belief`\" )))[i] - 1 # now 0 = will not be replicated, 1 = will be replicated confidence.rating <- eval(parse(text = paste0(\"prediction.data$`\",study.order[j],\"_\",label,\"conf_1`\")))[i] # on a scale from 0 - 100 confidence.rating <- ifelse(replication.belief == 0, confidence.rating*-1, confidence.rating) # make confidence in replication failure negative confidence.rating <- confidence.rating / 200 + .5 # convert to 0-1 scale study <- j condition <- ifelse(label == \"des.\",\"DescriptionOnly\",\"DescriptionPlusStatistics\") replication.outcome <- replication.outcomes[j] replication.effectsize <- replication.effectsizes[j] ind.subject.data <- cbind(study,condition,understanding,replication.belief,confidence.rating, replication.outcome,replication.effectsize) subject.data <- rbind(subject.data,ind.subject.data) } subject <- i ind.lay.data <- cbind(subject,subject.data) lay.data <- rbind(lay.data,ind.lay.data) } rm(study.order,label,condition,subject.data,understanding,replication.belief,confidence.rating, i,j,study,replication.outcome,replication.effectsize,ind.subject.data,subject,ind.lay.data)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "3. Data exclusion based on set criteria criteria: 1. if participants failed the attention check (i.e., did not press 'NO' and 75% (range 7080 is allowed)) 2. if a study description is not understood, exlcude this study for this participant 3. if a study is not understood by > 50% of the participants, exclude this study 4. if a participant does not understand > 50% of the studies, exclude this participant [CODESPLIT] bogus.study <- 27 clean.data <- as.data.frame(lay.data, stringsAsFactors = FALSE) correct.range <- clean.data[clean.data$study==bogus.study,]$confidence.rating correct.range <- rep(correct.range >= .1 & correct.range <= .15, each = nstudies) # correct range NO and 70-80% --> .1-.15 on the confidence scale clean.data.1 <- clean.data[correct.range,] # apply 1. clean.data.2 <- clean.data.1[is.na(clean.data.1$understanding),] # apply 2. remove.studies <- which(table(clean.data.2$study)<nsubjects/2) # indicate which studies are understood by less than half of the people clean.data.3 <- clean.data.2[! clean.data.2$study %in% remove.studies,] # apply 3. remove.subjects <- which(table(clean.data.3$subject)<nstudies/2) # indicate which subjects understood less than half of the studies clean.data.4 <- clean.data.3[! clean.data.3$subject %in% remove.subjects,] # apply 4. full.data <- clean.data.4 full.data$understanding <- NULL # delete empty column", "target": 1, "target_options": ["no_match", "match"]}
{"input": "total_distance of roads surveyed [CODESPLIT] sum(unique(site_data_cleaned$Site_Length_m))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Bring in census data for each site. [CODESPLIT] census_data <- fread(\"StudyAreas/Demographic_Site_Data/PDB_2015_Tract.csv\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot Quantiles against one another. [CODESPLIT] p2 <- ggplot() + geom_point(aes(y = receipt_distance_quantiles, x = montecarlo_distance_quantiles)) + geom_smooth(aes(y = receipt_distance_quantiles, x = montecarlo_distance_quantiles), method = \"lm\", se = F) + scale_x_log10(breaks= 10^c(1:5)) + scale_y_log10() + theme_classic(base_size = 20) + labs(x = \"Human Trip Quantile Distance (m)\", y = \"Receipt Quantile Distance (m)\") + coord_equal() + geom_abline(intercept = 0, slope = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "join the trash bearing data and wind data into single frame for plotting. [CODESPLIT] CompleteDataCoordPlot <- CompleteDataWithGoogle %>% filter(!is.na(windxcoordmeanbearing) & !is.na(trashxcoordbearing)) %>% #Removes data where we don't have wind or dont have receipt direction from the analysis. dplyr::select(windxcoordmeanbearing, windycoordmeanbearing, row) %>% rename(x = windxcoordmeanbearing, y = windycoordmeanbearing) %>% bind_rows(CompleteDataWithGoogle %>% filter(!is.na(windxcoordmeanbearing) & !is.na(trashxcoordbearing)) %>% dplyr::select(trashycoordbearing, trashxcoordbearing, row) %>% rename(x = trashycoordbearing, y = trashxcoordbearing))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Check if Trial column exists, if not create it [CODESPLIT] if (!\"Trial\" %in% colnames(df)) { df <- cbind(Trial = NA, df) } trial_counter <- 1 for (i in seq_along(index_pairs$start)) { start_pos <- index_pairs$start[i] end_pos <- index_pairs$end[i]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "group level Pi/Pb for each set size on probability scale [CODESPLIT] PiIdx <- NULL for (j in 1:nsubj) PiIdx <- c(PiIdx, which(names(as.data.frame(mcmcChain)) == paste0(\"Pi[\", j, \",\", s, \"]\"))) PbIdx <- NULL for (j in 1:nsubj) PbIdx <- c(PbIdx, which(names(as.data.frame(mcmcChain)) == paste0(\"Pb[\", j, \",\", s, \"]\"))) meanPi[s,experiment,] <- rowMeans(mcmcChain[, PiIdx]) # means across subjects of posterior means of Pi for each set size meanPb[s,experiment,] <- rowMeans(mcmcChain[, PbIdx]) } DeltaPi[,experiment] <- mcmcChain[, paste0(\"dPi\")] # posterior samples of slope of Pi over set size DeltaPb[,experiment] <- mcmcChain[, paste0(\"dPb\")] print(colMeans(DeltaPi>0)) # proportion of posterior samples > 0 save(MuPi, MuPb, meanPi, meanPb, DeltaPi, DeltaPb, file=parChainFile) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "determine manifestations and visualize cultural value dimensions [CODESPLIT] d$SVS_d1 <- d$SVS_harmony - d$SVS_mastery d$SVS_d2 <- d$SVS_egalit - d$SVS_hierarchy d$SVS_d3 <- (d$SVS_autona+d$SVS_autoni)/2 - d$SVS_embed cor(d[grepl(\"_d\",names(d))]) f$SVS_d1 <- f$SVS_harmony - f$SVS_mastery f$SVS_d2 <- f$SVS_egalit - f$SVS_hierarchy f$SVS_d3 <- (f$SVS_autona+f$SVS_autoni)/2 - f$SVS_embed cor(f[grepl(\"_d\",names(f))]) pcf <- princomp(f[,3:9], cor=T) summary(pcf) loadings(pcf)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "CTRees analyze the trees for each agreement indicator individually: Variationcorrelation of reported likelihood ~ proportion unstable: [CODESPLIT] png(\"output/figures/paper/ctree_rhopsi.png\", width = 600, height = 350) VS_ <- VS VS_ <- VS_[!is.na(VS_$rho_llhd) & !is.na(VS_$lagZ_20), ] VS_ <- VS_[!VS_$gtype_class %in% c(\"IFrc\", \"IFsc\", \"MFcr\", \"FCxr\") & VS_$band == \"TL\", ] VS_$gtype_class[VS_$gtype_class %in% \"SH\" & VS_$gtype_rank == \"secondary\"] <- \"SH/FC\" CT <- partykit::ctree(rho_llhd ~ nPDays , data = VS_, alpha = 0.05, maxdepth = 2) Vars <- setVars4ctreeColoring(VS_) plotCTree(CT) dev.off()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Selecting and recoding items that were measured in all waves [CODESPLIT] d <- df %>% select(id, wave, lsat = sat6, fsat = sat1i3, per1i2, per1i7, per1i13) %>% mutate(per1i2 = invert(per1i2, 5))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "6. Extract & plot estimates Is evolution more comparable to B0 or B1? [CODESPLIT] describe_posterior((r.delta.z[,1,])) # B0_Conc describe_posterior((r.delta.z[,2,])) # B0_Disc describe_posterior((r.delta.z[,3,])) # B1_Conc describe_posterior((r.delta.z[,4,])) # B1_Disc", "target": 1, "target_options": ["no_match", "match"]}
{"input": "for each subject, aggregate his/her selfratings across all interactions (in second and third week of eventbased assessment) [CODESPLIT] app_SR_aggr <- aggregate(app_SR, by=list(app_SR$id_a), mean)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "recode abitur grades which were not provided but stored as 0, and grades that were provided in units of the wrong grading system [CODESPLIT] surv2345$abitur_grade[surv2345$abitur_grade == 0] <- NA surv2345$abitur_grade[surv2345$abitur_grade == 12] <- 2.0 surv2345$abitur_grade[surv2345$abitur_grade == 13] <- 1.7", "target": 1, "target_options": ["no_match", "match"]}
{"input": "define a function which, for each person, selects the first nonNA measurement he provided out of three time points ( out of three variables) [CODESPLIT] firstnonNA_3timepoints <- function(df) {if(all(is.na(df))){ NA } else if (!is.na(df[1])){ df[1] } else if (!is.na(df[2])){ df[2] } else if (!is.na(df[3])){ df[3] } }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "for the exam grades 15 to 18 which have been assessed only at the last occasions (Survey 5), save the grade in new variable for consistency in notation [CODESPLIT] surv2345$retro_grades_15 <- surv2345$retro_grades_15_t5 surv2345$retro_grades_16 <- surv2345$retro_grades_16_t5 surv2345$retro_grades_17 <- surv2345$retro_grades_17_t5 surv2345$retro_grades_18 <- surv2345$retro_grades_18_t5", "target": 1, "target_options": ["no_match", "match"]}
{"input": "select variables (all 18 exam grades) to be aggregated for mean exam grade [CODESPLIT] retro_grades_df <- as.matrix(surv2345[,paste0(\"retro_grades_\",1:18)])", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Select variables for the analyses and save data frame that will be uploaded in the OSF [CODESPLIT] connect_osf <- select(connect, Z_Raven_self:Z_MWTB_obj, Z_global_selfeval:Z_achievement) write.table(connect_osf, file=\"Data_Sample_C_connect.txt\",sep = \"\\t\",col.names=TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "DESCRIPTIVE STATISTICS compute and save sample statistics (age distribution, number of females) [CODESPLIT] age <- round(select(psych::describe(connect_descr$age), n, min, max, mean, sd),2) age$n <- nrow(connect_descr) sampstats <- mutate(age, female=plyr::count(connect_descr$sex)[plyr::count(connect_descr$sex)[,1]==\"1\",][\"freq\"] ) write.table(sampstats, file=\"Descriptives/age_sex_Sample_C_connect.dat\", sep=\"\\t\", row.names=FALSE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compute and save correlation table of variables before aggregation and standardization [CODESPLIT] cor_raw <- corcons(connect_descr) write.table(cor_raw, file=\"Descriptives/correlations_raw_Sample_C_connect.dat\", sep=\"\\t\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hypothesis 1.1 Hypothesis 1 was tested via pairedsample t.tests (alphas need to be Bonferronicorrected) Examples for conscientious goal classes (first two rows of Table 1 Conscientious Goals). Remaining tests are simialr [CODESPLIT] t.test(x = dt1$CP_Class_01, y = dt1$CN_Class_01, paired = TRUE, alternative = \"greater\") t.test(x = dt1$CP_Class_02, y = dt1$CN_Class_02, paired = TRUE, alternative = \"greater\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hypothesis 1.2 Hypothesis 1.2 was tested via partiallyoverlapping samples ttests. We report the analyses for the first panel of Table S2 (and for pvalues reported in Table 2). The remaining analyses are identical, but were performed on other goal classes [CODESPLIT] Partover.test(dt1$CP_Class_01, dt1$HP_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$EP_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$XP_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$AP_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$OP_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$HN_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$EN_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$XN_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$AN_Class_01, stacked = TRUE, alternative = \"greater\") Partover.test(dt1$CP_Class_01, dt1$ON_Class_01, stacked = TRUE, alternative = \"greater\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Examples of Tobit regressions of the subjective importance of specific goals on HEXACO traits (first two lines of Table 3). Pvalues need to be corrected using the Bonferroni method, considering that we performed 21 multiple regressions (multiply them by 21 and round values larger than 1 to 1) [CODESPLIT] predictors <- select(dt2, HEXACO_H:HEXACO_O) %>% scale() fit1 <- censReg(dt2$G_C_01_DimostrareQlcQln_Importance ~ predictors, left = 1, right = 9) summary(fit1) fit2 <- censReg(dt2$G_C_02_EssereDegnoFiducia_Importance ~ predictors, left = 1, right = 9) summary(fit2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hierarchical regression predicting the willingness to change conscientiousness according to CBFI (Table 4). For ease of formatting, I used the R package AutoModel [CODESPLIT] Data4HierarchicalReg <- select(dt2, CBFI_C, POS, HEXACO_H:HEXACO_O, BFI2_O:BFI2_N, Importance_GC, Importance_GU) %>% scale() %>% data.frame() Data4HierarchicalReg$BFTGI_C_bin <- as.numeric(dt2$BFTGI_C == 3)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Hypothesis 3.2 multiple regressions predicting each goal class from HEXACO traits (we report examples reproducing the first two rows of Table 8, the code for the others is similar, save for the goal class). Pvalues need to be corrected using the Bonferroni method, considering 9 multiple regressions (multiply them by 9 and transform values > 1 to 1) [CODESPLIT] lm(G08Rules ~ ., data = select(dt3, G08Rules, HEXACO_H:HEXACO_O)) %>% lm.beta %>% summary lm(G10Control ~ ., data = select(dt3, G10Control, HEXACO_H:HEXACO_O)) %>% lm.beta %>% summary", "target": 1, "target_options": ["no_match", "match"]}
{"input": "DIFFERENTIAL ITEM FUNCTIONING Create trichotomous income variable [CODESPLIT] data = data %>% mutate( income3 = recode(income, '1=1;; 2=1;; 3=1;; 4=1;; 5=1;; 6=1;; 7=1;; 8=1;; 9=2;; 10=2;; 11=2;; 12=2;; 13=2;; 14=3;; 15=3;; 16=3;; 17=3;; 18=3')) q1merit = data[c(\"Q1A\", \"Q1B\", \"Q1H\", \"Q1N\")] q1opportunity = data[c(\"Q1C\", \"Q1D\", \"Q1F\", \"Q1G\")] q1chance = data[c(\"Q1I\", \"Q1O\", \"Q1P\", \"Q1Q\")] q2merit = data[c(\"Q2A\", \"Q2B\", \"Q2H\", \"Q2N\")] q2opportunity = data[c(\"Q2C\", \"Q2D\", \"Q2F\", \"Q2G\")] q2chance = data[c(\"Q2I\", \"Q2O\", \"Q2Q\", \"Q2R\")] q3merit = data[c(\"Q3A\", \"Q3B\", \"Q3H\", \"Q3N\")] q3opportunity = data[c(\"Q3C\", \"Q3D\", \"Q3F\", \"Q3G\")] q3chance = data[c(\"Q3I\", \"Q3O\", \"Q3P\", \"Q3Q\")] q4merit = data[c(\"Q4B\", \"Q4C\", \"Q4I\", \"Q4N\")] q4opportunity = data[c(\"Q4D\", \"Q4E\", \"Q4G\", \"Q4H\")] q4chance = data[c(\"Q4J\", \"Q4O\", \"Q4P\", \"Q4Q\")] sesmerit = data[c(\"hs_merit\", \"ls_merit\", \"ha_merit\", \"la_merit\")] sesopportunity = data[c(\"hs_opportunity\", \"ls_opportunity\", \"ha_opportunity\", \"la_opportunity\")] seschance = data[c(\"hs_chance\", \"ls_chance\", \"ha_chance\", \"la_chance\")] merit_full = data[c(\"hs_efft\", \"hs_abil\", \"ls_efft\", \"ls_abil\", \"ha_efft\", \"la_efft\")] gender = data$gender age = data$AGE4 politic = data$D3 income = data$income3 educ = data$EDUC4", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Display distribution of efficacy judgments (can be used to set a prior on the next experiment) [CODESPLIT] eff_all <- c(eff_nograph, eff_graph) par(mfrow=c(2,3)) myhist(\"Efficacy\", eff_all, 1, 9, N/2) myhist(\"Efficacy - no graph\", eff_nograph, 1, 9, N/4) myhist(\"Efficacy - graph\", eff_graph, 1, 9, N/4) cat(\"Grand mean for efficacy: \", format_number(mean(eff_all)), \"\\n\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Models by social network measure 1. Indegree [CODESPLIT] m1 <- glmer(pup_year_surv~indegree + valley + overall.index + pup_sex + pup_littersizeborn + mother_age + network_size + pup_emerjdate + overall.index*indegree + valley*indegree + (1|mother_uid) + (1|pup_yrborn), control = glmerControl(\"bobyqa\", optCtrl=list(maxfun=2e5)), data=sur_data, family= binomial) summary(m1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove upper triangle of correlation matrix [CODESPLIT] if(removeTriangle[1]==\"upper\"){ Rnew <- as.matrix(Rnew) Rnew[upper.tri(Rnew, diag = TRUE)] <- \"\" Rnew <- as.data.frame(Rnew) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "put SDs in parantheses [CODESPLIT] mutate(gender_sd = paste0(\"(\", gender_sd, \")\"), age_sd = paste0(\"(\", age_sd, \")\"), parent_sd = paste0(\"(\", parent_sd, \")\"), pol_sd = paste0(\"(\", pol_sd, \")\")) %>% mutate(gender_sd = paste0(\"(\", gender_sd, \")\"), age_sd = paste0(\"(\", age_sd, \")\"), parent_sd = paste0(\"(\", parent_sd, \")\"), pol_sd = paste0(\"(\", pol_sd, \")\")) %>% mutate(intent_apply_job_sd = paste0(\"(\", intent_apply_job_sd, \")\"), intent_apply_ps_sd = paste0(\"(\", intent_apply_ps_sd, \")\"), psm_sd = paste0(\"(\", psm_sd, \")\"), pofit_sd = paste0(\"(\", pofit_sd, \")\"), pjfit_sd = paste0(\"(\", pjfit_sd, \")\")) %>% mutate(intent_apply_job_sd = paste0(\"(\", intent_apply_job_sd, \")\"), intent_apply_ps_sd = paste0(\"(\", intent_apply_ps_sd, \")\"), psm_sd = paste0(\"(\", psm_sd, \")\"), pofit_sd = paste0(\"(\", pofit_sd, \")\"), pjfit_sd = paste0(\"(\", pjfit_sd, \")\")) %>%", "target": 1, "target_options": ["no_match", "match"]}
{"input": "5.3 Pirateplot for dependent variables [CODESPLIT] df_pirate_dv <- tibble(dv = c(df$intent_apply_job, df$intent_apply_ps), group = c(rep.int(0, length(df$intent_apply_job)), rep.int(1, length(df$intent_apply_ps)))) %>% mutate(group = factor(group, labels = c(\"Intention to\\napply for job\", \"Intention to apply\\nfor public service\"))) pirate_dvs <- yarrr::pirateplot(dv ~ group, data = df_pirate_dv, inf.method = \"ci\", xlab = \"Dependent variable\", ylab = \"\", theme = 2, cex.lab = 1.2, cex.axis = 1.2, cex.names = 1.2) pirate_dvs <- recordPlot() # contains all plotting information png(\"./output/Appendix_pirate_dvs.png\") pirate_dvs dev.off()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Load the functions for the raincloud plots [CODESPLIT] source('funcs/R_rainclouds.R')", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Save the result and load if you run the permutation test on a different computer than the visualizations [CODESPLIT] save(list.res,file=\"results.Rdata\") load(\"results.Rdata\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "data as a scatter plot, column s for x and column v for y, size 4 points, use the sat_curve data frame as the source [CODESPLIT] geom_point(data = sat_curve, aes(x = s, y = v), size = 4) +", "target": 1, "target_options": ["no_match", "match"]}
{"input": "fit as a line plot, use the mm_fit data frame as the data source [CODESPLIT] geom_line(data = mm_fit, aes(x = s, y = v)) +", "target": 1, "target_options": ["no_match", "match"]}
{"input": "change the axis label text to size 22, bold, and black color [CODESPLIT] axis.text.x = element_text(size = 22, face = \"bold\", color = \"black\"), axis.text.y = element_text(size = 22, face = \"bold\", color = \"black\")) +", "target": 1, "target_options": ["no_match", "match"]}
{"input": "sequence of age inputs within range of model: (15, 75) years old [CODESPLIT] seq.length <- length(seq(15, 75, 1))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Define link function with softmax transformation The following function is taken from Koster and McElreath (2017) Modifications made to accommodate current model structure (e.g. addition of com_id, month) [CODESPLIT] link.mn <- function( data ) { K <- dim(post$v_id)[3] + 1 ns <- dim(post$v_id)[1] if ( missing(data) ) stop( \"BOOM: Need data argument\" ) n <- seq.length softmax2 <- function(x) { x <- max(x) - x exp(-x)/sum(exp(-x)) } p <- list() for ( i in 1:n ) { p[[i]] <- sapply( 1:K , function(k) { if ( k < K ) { ptemp <- post$a[,k] + post$bA[,k] * data$age_z[i] + post$bQ[,k] * data$age_zq[i] + post$bT[,k] * data$time_z[i] + post$bTQ[,k] * data$time_zq[i] if ( data$id[i]>0 ) ptemp <- ptemp + post$v_id[,data$id[i],k] if ( data$com_id[i]>0 ) ptemp <- ptemp + post$v_com[,data$com_id[i],k] if ( data$month_id[i]>0 ) ptemp <- ptemp + post$v_month[,data$month_id[i],k] } else { ptemp <- rep(0,ns) } return(ptemp) })", "target": 1, "target_options": ["no_match", "match"]}
{"input": "The values are converted to probabilities using the softmax function which ensures that the predicted values across categories sum to 100% probabilities. [CODESPLIT] for ( s in 1:ns ) p[[i]][s,] <- softmax2( p[[i]][s,] ) } return(p) } for ( s in 1:ns ) p[[i]][s,] <- softmax2( p[[i]][s,] ) } return(p) } low_age <- (15 - mean_age_female)/sd_age_female high_age <- (75 - mean_age_female)/sd_age_female", "target": 1, "target_options": ["no_match", "match"]}
{"input": "everything will be calculated over the \"adult\" interval from 15 to 75 [CODESPLIT] age_seq <- (seq(15, 75, 1)- mean_age_female)/sd_age_female", "target": 1, "target_options": ["no_match", "match"]}
{"input": "We create a data frame from age_seq and its second order polynomial, holding most of predictors at their sample mean. In the original Koster+McElreath paper, they specified 8:00 am for the time of day in order to effect a more sensible alignment with the empirical data. In this case, we hold that parameter at 0, which therefore corresponds with the mean time of day in our dataset for behavioral observations (0.532 ~12:45). Also, as we noted earlier in the notes about the multinomial link function, we set \"id\" to zero in order to average over the random effects. [CODESPLIT] pred_dat <- data.frame( id = 0 , age_z = age_seq, age_zq = age_seq^2, time_z = 0, time_zq = 0, com_id = 0, month_id = 0 ) p <- link.mn ( pred_dat )", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Biodiversity conversion general form is Index (Maxent ^ wM + Abundance ^ wL )^(1/(wM+wL)) but would need to take care of how this scales [CODESPLIT] pro.fun <- function(x, abuntab, zonename){ (x^wM*abuntab[,which(names(abuntab)==zonename)]^wL)^(1/(wM+wL))}", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Construct variable for number of family members in country [CODESPLIT] famcountry = data4$partnerwhere + data4$mumwhere + data4$dadwhere table(famcountry) data4$famcountry <- famcountry hist(data4$famcountry) sum(is.na(data4$famcountry)) #175", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot network latent variable on CFA latent variable [CODESPLIT] layout(t(c(1,2))) plot(cfaLV, netLV, main = \"Network Latent Variable\\non CFA Latent Variable\", ylab = \"Network Latent Variable\", xlab = \"CFA Latent Variable\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot participant means on CFA latent variable [CODESPLIT] plot(cfaLV, pmeans, main = \"Participant Means on\\nCFA Latent Variable\", ylab = \"Participant Means\", xlab = \"CFA Latent Variable\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "initialize matrices to store validation results [CODESPLIT] WklValidation <- matrix(nrow = nrow_max, ncol = 13, dimnames = list(seq(nrow_max), c(\"distribution\", \"nProfiles_checked\", \"day_after_burial\", \"timewindow_lower\", \"timewindow_upper\", \"likelihood_reported\", \"likelihoodSpread_reported\", \"distribution_reported\", \"sensitivity_reported\", \"size_reported\", \"sizeSpread_reported\", \"grain_size_reported\", \"data_quality\"))) WklValidation_char <- matrix(nrow = nrow_max, ncol = 10, dimnames = list(seq(nrow_max), c(\"vf_uuid\", \"pwl_uuid\", \"pwl\", \"validation_date\", \"band\", \"timing_mode\", \"gtype_class\", \"gtype_rank\", \"comment\", \"danger_rating\"))) vf_uuid <- NA", "target": 1, "target_options": ["no_match", "match"]}
{"input": "run linear models on predicted probabilities, to calculate the trend [CODESPLIT] models <- pred %>% group_by(x, country) %>% nest() %>% mutate(models = map(data, ~ lm(predicted ~ wave, data = .))) %>% spread_coef(models, se = TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "one participant has a typing error in the age variable (stated he was 2). Set to NA. [CODESPLIT] mst1[which(mst1$age == 2), \"age\"] <- NA", "target": 1, "target_options": ["no_match", "match"]}
{"input": "association between vote and ideology [CODESPLIT] ggplot(dat, aes(x = Ideology, y = Percent.Vote.Rep)) + geom_vline(xintercept = .50, color = \"gray50\", linetype = 2) + geom_hline(yintercept = .50, color = \"gray50\", linetype = 2) + stat_poly_line() + stat_correlation(label.x = \"left\", small.r = TRUE) + geom_point() + geom_text_repel(size = 3, aes(label = Group)) + labs(y = \"Proportion Voting Republican\", x = \"Ideology\") + coord_cartesian(xlim = c(0, 1), ylim = c(0, 1)) + theme_cowplot() ggsave(\"figures/predictor_scatter.pdf\", width = 5.5, height = 5)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Make a dot and a line color column depending on the significance level Check significance level and if significant add respective color to new column [CODESPLIT] ageR = data.frame(dotCol = rep(ifelse(summary(lm(initLevel ~ age, data = id_df))$coefficients[2,4]< 0.05, \"khaki3\", \"lightgrey\"), length(unique(id_df$userCode))), lineCol = rep(ifelse(summary(lm(initLevel ~ age, data = id_df))$coefficients[2,4]< 0.05, \"cyan4\", \"darkgrey\"), length(unique(id_df$userCode)))) educR = data.frame(dotCol = rep(ifelse(summary(lm(initLevel ~ educ_y, data = id_df))$coefficients[2,4]< 0.05, \"khaki3\", \"lightgrey\"), length(unique(id_df$userCode))), lineCol = rep(ifelse(summary(lm(initLevel ~ educ_y, data = id_df))$coefficients[2,4]< 0.05, \"cyan4\", \"darkgrey\"), length(unique(id_df$userCode)))) blpR = data.frame(dotCol = rep(ifelse(summary(lm(initLevel ~ BLP, data = id_df))$coefficients[2,4]< 0.05, \"khaki3\", \"lightgrey\"), length(unique(id_df$userCode))), lineCol = rep(ifelse(summary(lm(initLevel ~ BLP, data = id_df))$coefficients[2,4]< 0.05, \"cyan4\", \"darkgrey\"), length(unique(id_df$userCode)))) actR = data.frame(dotCol = rep(ifelse(summary(lm(initLevel ~ Act_total, data = id_df))$coefficients[2,4]< 0.05, \"khaki3\", \"lightgrey\"), length(unique(id_df$userCode))), lineCol = rep(ifelse(summary(lm(initLevel ~ Act_total, data = id_df))$coefficients[2,4]< 0.05, \"cyan4\", \"darkgrey\"), length(unique(id_df$userCode)))) motivR = data.frame(dotCol = rep(ifelse(summary(lm(initLevel ~ socioAffect_init, data = id_df))$coefficients[2,4]< 0.05, \"khaki3\", \"lightgrey\"), length(unique(id_df$userCode))), lineCol = rep(ifelse(summary(lm(initLevel ~ socioAffect_init, data = id_df))$coefficients[2,4]< 0.05, \"cyan4\", \"darkgrey\"), length(unique(id_df$userCode)))) tmp = rbind.data.frame(ageR, educR, blpR, actR, motivR) corrDF = cbind(corrDF, tmp)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Sample from those names 200 times, with replacement [CODESPLIT] animal <- sample(categories, 200, replace = TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create random variables Use rnorm() with varying means and SDs [CODESPLIT] length <- rnorm(200, 30, 8) weight <- rnorm(200, 4, 2) happiness <- rnorm(200, 3.5, 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Examine descriptive statistics Use the pipe, group_by(), and summarise() to get means, SDs, and medians for each category [CODESPLIT] animal_data %>% group_by(animal) %>% summarise( weight_mean = mean(weight), weight_sd = sd(weight), weight_median = median(weight) )", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Subset data to include only the first two categories Use filter() to select only rows that pass the given logical test [CODESPLIT] subset_1 <- animal_data %>% filter(animal == \"puppy\" | animal == \"kitten\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Use subset for ttest Welch's ttest [CODESPLIT] t.test(weight ~ animal, data = subset_1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot spaghettiplot for minute 1 and minute 5 [CODESPLIT] df_1_5 <- rbind(df_min1, df_min5) df_1_5 <- df_1_5 %>% mutate(range=droplevels(range)) p <- ggplot(data = df_1_5, aes(x = range, y = count, group = pp)) p + geom_point() p + geom_line() p + geom_line() + facet_grid(. ~ condition) p <- ggplot(data = df_min, aes(x = range, y = count, group = pp)) p + geom_point() p + geom_line() p + geom_line() + facet_grid(. ~ condition)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "do 3 way anova including the factor in which condition they are in [CODESPLIT] bfFull = anovaBF(count ~ condition + range + experiment + condition:range:experiment, data = df_b) bfFull[4]/bfFull[3] bfFull[18]/bfFull[17] bf1 = anovaBF(count ~ condition + range + experiment, data = df_b)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "pc: critical pvalue Overview: Creates a vector of the same length as the number of tests submitted to pcurve, significant and not, and computes the proportion of pvalues expected to be smaller than {pc} given the d.f. and outputs the entire vector, with NA values where needed Ftests (& thus ttests) [CODESPLIT] prop=ifelse(family==\"f\" & p<.05,1-pf(qf(1-pc,df1=df1, df2=df2),df1=df1, df2=df2, ncp=ncp33),NA)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create vector that numbers studies 1 to N,includes n.s. studies [CODESPLIT] k=seq(from=1,to=length(raw))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "1.2 Parse the entered text into usable statistical results 1.3 Create test type indicator [CODESPLIT] stat=substring(raw,1,1) #stat: t,f,z,c,r test=ifelse(stat==\"r\",\"t\",stat) #test: t,f,z,c (r-->t)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Make red dot at the estimate [CODESPLIT] points(hat,min(fit,na.rm=TRUE),pch=19,col=\"red\",cex=2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "This loop creates a vector, blue, with 5 elements, with the proportions of p.01,p.02...p.05 [CODESPLIT] for (i in c(.01,.02,.03,.04,.05)) blue=c(blue,sum(ps==i,na.rm=TRUE)/ksig*100)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "FIG 3: scree plot for parallel analysis [CODESPLIT] rscree<-as.data.frame(cbind(spar$fa.values, spar$fa.simr)) colnames(rscree)<-c(\"Actual\",\"Resampled\") rscree$item<-seq.int(1, 12, 1) refplot<-pivot_longer(rscree, cols=1:2, names_to=\"Method\") refplot2<-refplot %>% filter(item<5) iscree<-as.data.frame(cbind(ipar$fa.values, ipar$fa.simr)) colnames(iscree)<-c(\"Actual\",\"Resampled\") iscree$item<-seq.int(1, 8, 1) insplot<-pivot_longer(iscree, cols=1:2, names_to=\"Method\") insplot2<-insplot %>% filter(item<5) refeig<-ggline(refplot2, x=\"item\", y=\"value\", group = \"Method\", color=\"Method\", size=1.1, palette = paletteer_d(\"palettetown::tangela\", direction=-1), xlab=\"Factor\", ylab=\"Eigenvalue\", title=\"Self-reflection\")+ theme_minimal_hgrid(font_size=12, font_family = \"Fira Sans Medium\") scree.a<-ggpar(refeig, ylim = c(0,6), yticks.by=1, legend = c(.7,.77), legend.title = \"\") scree.a inseig<-ggline(insplot2, x=\"item\", y=\"value\", group = \"Method\", color=\"Method\", size=1.1, palette = paletteer_d(\"palettetown::tangela\", direction=-1), xlab=\"Factor\", ylab=\"\", title=\"Insight\")+ theme_minimal_hgrid(font_size=12, font_family = \"Fira Sans Medium\") scree.b<-ggpar(inseig, ylim = c(0, 6), yticks.by=1, legend = c(.7,.77), legend.title = \"\") scree.b", "target": 1, "target_options": ["no_match", "match"]}
{"input": "make sure data are in chronological order [CODESPLIT] pollenData <- pollenData[order(pollenData$age),] lakeData <- lakeData[order(lakeData$age),] charcoalData <- charcoalData[order(charcoalData$age),] tail(pollenData) tail(lakeData) tail(charcoalData)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "examine descriptives of newly created variables here we are combining the dplyr and summarytools packages to subset the data and then get descriptives [CODESPLIT] eid_dat %>% dplyr::select(swl_mean, meim_ex_mean, meim_co_mean) %>% summarytools::descr()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "step 6: multiple regression with exploration and commitment predicting SWL this function is in base R so no package to call [CODESPLIT] regression <- lm(swl_mean ~ meim_ex_mean + meim_co_mean, data=eid_dat) summary(regression) confint(regression)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "step 7: create scattterplots using ggplot2 scatterplot of exploration with SWL there are many more options you can play with for customization. I just included some basics here [CODESPLIT] ex_swl_scatter <- ggplot2::ggplot(eid_dat, aes(meim_ex_mean, swl_mean)) + geom_point() + ggtitle(\"My Scatterplot\") + xlab(\"Ethnic Identity Exploration (mean)\") + ylab(\"Satisfaction with Life (mean)\") + geom_smooth(method=\"lm\") ex_swl_scatter", "target": 1, "target_options": ["no_match", "match"]}
{"input": "scatterplot of commitment with SWL [CODESPLIT] co_swl_scatter <- ggplot2::ggplot(eid_dat, aes(meim_co_mean, swl_mean)) + geom_point() + ggtitle(\"My Scatterplot\") + xlab(\"Ethnic Identity Commitment (mean)\") + ylab(\"Satisfaction with Life (mean)\") + geom_smooth(method=\"lm\") co_swl_scatter", "target": 1, "target_options": ["no_match", "match"]}
{"input": "step 8: indepedent samples ttests also in base R. note that Welch's ttest is the default this set is testing for mean differences by whether they were born in the U.S. [CODESPLIT] t.test(eid_dat$meim_ex_mean ~ eid_dat$usborn) t.test(eid_dat$meim_co_mean ~ eid_dat$usborn) t.test(eid_dat$swl_mean ~ eid_dat$usborn)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "step 11: create a series of boxplots to go with the previous ANOVAs note that the code here is only slightly different from the one variable case you can/should add elements to format the plots, as you did previously [CODESPLIT] borngen_ex_boxplot <- ggplot2::ggplot(eid_dat, aes(usborn, meim_ex_mean, fill=firstgen)) + geom_boxplot(alpha = 0.5) borngen_ex_boxplot borngen_co_boxplot <- ggplot2::ggplot(eid_dat, aes(usborn, meim_co_mean, fill=firstgen)) + geom_boxplot(alpha = 0.5) borngen_co_boxplot borngen_swl_boxplot <- ggplot2::ggplot(eid_dat, aes(usborn, swl_mean, fill=firstgen)) + geom_boxplot(alpha = 0.5) borngen_swl_boxplot", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add a plotting function to get power curve [CODESPLIT] plot_power_extend <- function(object, target = .80, ...) { nsim <- object[[1]]$n df <- purrr::map_dfr(object, powerinterval, .id = \"n\", ...) df$n <- as.integer(df$n) ggplot(df, aes(x = n, y = mean, group = 1)) + geom_hline(yintercept = target) + geom_smooth(method = \"glm\", formula = cbind(y * nsim, (1 - y) * nsim) ~ x, method.args = list(family = binomial(\"probit\"))) + geom_pointrange(aes(ymin = lower, ymax = upper)) + labs(y = \"power\") + ylim(0, 1) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check duplicate trials per participant (post) [CODESPLIT] xtabs(~subj+trial,data=rawdat.all) table(rawdat.all$subj)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove duplicate species per file, and species level data [CODESPLIT] data2.5 <- data2[!data2$species == 'Certhia_sp.',] data2.6 <- data2.5[!data2.5$species == 'Regulus_sp.',] data3 <- data2.6 %>% group_by(Plot_no) %>% mutate(whichday = as.integer(factor(date))) data4 <- data3 %>% group_by(Plot_no, filename) %>% mutate(seq=cur_group_id())", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Show sample sizes per country [CODESPLIT] pa12_resp_s %>% group_by(CNT) %>% summarise(n = n())", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Gaze Duration Descriptive stats per ID [CODESPLIT] BFSGDesc <- BirdFSG %>% group_by(ID, Stimuli) %>% summarise (MGaze = mean(Gaze), sd=sd(Gaze), n=n(), se = sd/sqrt(n)) ;; BFSGDesc$Stimuli <- paste( BFSGDesc$Stimuli, \"S\", sep=\"\") BFSGDesc BNADDesc <- BirdNAD %>% group_by(ID, Stimuli) %>% summarise (MGaze = mean(Gaze), sd=sd(Gaze), n=n(), se = sd/sqrt(n)) ;; BNADDesc$Stimuli <- paste( BNADDesc$Stimuli, \"S\", sep=\"\") BFSGDesc <- as.data.frame(BFSGDesc);; BNADDesc <- as.data.frame(BNADDesc);; BFSGDesc$Stimuli <- as.factor(as.character(BFSGDesc$Stimuli)) BNADDesc$Stimuli <- as.factor(as.character(BNADDesc$Stimuli)) BFSGDesc", "target": 1, "target_options": ["no_match", "match"]}
{"input": "p<.05 ONEWAY ANOVA let's say we're interested in investigating the relationship between three species (setosa, versicolor, virginica) outcome: sepal width run Levene's test to check assumption of homogeneity of variance by default centres variable using median library(car) leveneTest(Sepal.Width~Species, datairis,centermean) p>.05 therefore assumption met running oneway ANOVA [CODESPLIT] mod.a<-aov(Sepal.Width~Species, data=iris) summary(mod.a)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "generate a random number between 1 and 1000 that functions as the index into an array of random numbers of size 1000 [CODESPLIT] return(randNo[floor(runif(1)*1000)]) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "print the simulation number after every 10% of the simulations [CODESPLIT] if(s%%(sims/10) == 0){ cat(paste(\" \", s)) } if(verbose){ print(\"\",quote=FALSE) print(\"\",quote=FALSE) print(\"=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-\", quote=FALSE) print(paste(\" Simulation\", s, \" \"), quote=FALSE) print(\"=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-\", quote=FALSE) } for(i in c(1:stages)){ if(i==picCreStage){ add.time(par.dat) if(verbose){ print(paste(\"------ Adding picture chunks -------\"), quote=FALSE) } p.knopf[[\"ctime\"]] = currTime add.chunk(p.knopf) p.flasche[[\"ctime\"]] = currTime add.chunk(p.flasche) if(EXPT1){ p.ballon[[\"ctime\"]] = currTime add.chunk(p.ballon) p.blume[[\"ctime\"]] = currTime add.chunk(p.blume) } martin[[\"ctime\"]] = currTime add.chunk(martin) sarah[[\"ctime\"]] = currTime add.chunk(sarah) } if(verbose){ print(\"\",quote=FALSE) print(paste(\"=============== \", \"Stage \", i, \":\", input.stages[i], \" ===============\"), quote=FALSE) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "mean level of transitive respondents per wave [CODESPLIT] aggregate(long_svo_multi$trans,list(long_svo_multi$wave_cat),FUN=mean,na.rm =T)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create variable with only NA values [CODESPLIT] long_svo_multi$vlengths = rep(NA, times = 2970)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate vector length per row [CODESPLIT] for (i in 1:2970) {long_svo_multi$vlengths[i]=calculate_vlength(mean_self[i], mean_other[i])}", "target": 1, "target_options": ["no_match", "match"]}
{"input": "criterion to select applicable cases select only rows with transitive and good vector length response profiles as well as respondents with NA values this criterion used to select applicable cases for data analyses and figures [CODESPLIT] data_criterion <- (long_svo_multi$trans == \"TRUE\" & long_svo_multi$trans_vlenght == \"TRUE\" | is.na(long_svo_multi$trans) | is.na(long_svo_multi$trans_vlenght))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "KolmogorovSmirnov ksamples test for SVO angles [CODESPLIT] ks.test(dropouts$SVO_angle, stayers$SVO_angle) ks.test(dropouts$SVO_angle, stayers$SVO_angle) ks.test(dropouts$SVO_angle, stayers$SVO_angle) ks.test(dropouts$SVO_angle, stayers$SVO_angle) ks.test(dropouts$SVO_angle, stayers$SVO_angle)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fishersexact test for SVO types [CODESPLIT] fisher.test(table(dropouts$SVO_dicho_num),table( stayers$SVO_dicho_num)) fisher.test(table(dropouts$SVO_dicho_num),table( stayers$SVO_dicho_num)) fisher.test(table(dropouts$SVO_dicho_num),table( stayers$SVO_dicho_num)) fisher.test(table(dropouts$SVO_dicho_num),table( stayers$SVO_dicho_num)) fisher.test(table(dropouts$SVO_dicho_num),table( stayers$SVO_dicho_num))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "lagged variables of distance to 45 boundary in SVO [CODESPLIT] long_svo_multi$distance_45<-round(abs(45 - long_svo_multi$prior_SVO),digits = 2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "alter a variable to label dropouts as 1 and stayers as 0 for ML analysis and this variable as dependent variable [CODESPLIT] data_adjusted$stay_dropout <- data_adjusted$change_cat data_adjusted$stay_dropout <- ifelse(data_adjusted$stay_dropout == 1,0, ifelse(data_adjusted$stay_dropout == 0,0,1)) data_adjusted$stay_dropout <- replace_na(data_adjusted$stay_dropout,value = 1) data_adjusted_1 <- data_adjusted %>% filter(trans == \"TRUE\" & trans_vlenght == \"TRUE\" | is.na(trans) | is.na(trans_vlenght))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "grid arrange to create figure with distribution of angles and SVO categories [CODESPLIT] get_legend<-function(myggplot){ tmp <- ggplot_gtable(ggplot_build(myggplot)) leg <- which(sapply(tmp$grobs, function(x) x$name) == \"guide-box\") legend <- tmp$grobs[[leg]] return(legend)} bp <- theme(legend.position=\"none\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Assessing minor longitudinal differences in SVO angles select applicable cases and variables [CODESPLIT] adjust <-dplyr::select(long_svo_multi,id,SVO_angle,wave_cat,trans,trans_vlenght) data_input <- na.omit(adjust) data_input$SVO_angle <- round(data_input$SVO_angle, digits = 2) data_input <- data_input %>% filter(trans == \"TRUE\" & trans_vlenght == \"TRUE\") data_wide_2 <- dcast(data_input, id ~ wave_cat, value.var=\"SVO_angle\") data_wide_2 <- na.omit(data_wide_2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "cumulative change of difference in SVO angles wave per wave [CODESPLIT] data_wide_6$change_wave_1 <- data_wide_6$diff_w1 data_wide_6$change_wave_1_2 <- data_wide_6$diff_w1_w2 data_wide_6$change_wave_1_2_3 <- data_wide_6$change_wave_1_2 + data_wide_6$diff_w2_w3 data_wide_6$change_wave_1_2_3_4 <- data_wide_6$change_wave_1_2_3 + data_wide_6$diff_w3_w4 data_wide_6$change_wave_1_2_3_4_5 <- data_wide_6$change_wave_1_2_3_4 + data_wide_6$diff_w4_w5 data_wide_6$change_wave_1_2_3_4_5_6 <- data_wide_6$change_wave_1_2_3_4_5 + data_wide_6$diff_w5_w6 data_wide_7 <- data_wide_6[, c(1,8:13)] long_SVO_4 <- melt(data = data_wide_7, id.vars = c(\"id\"), variable.name = \"wave_compare\", value.name = \"cumulative_change_in_SVO\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Set up ERGM formula and constraints [CODESPLIT] if (substr(thresh, 1, 4) == 'prop') { ergm_formula <- y ~ gwesp(0.75, fixed=TRUE) + gwnsp(0.75, fixed=TRUE) constraints <- ~edges } else { ergm_formula <- y ~ edges + gwesp(0.75, fixed=TRUE) + gwnsp(0.75, fixed=TRUE) constraints <- ~. } ergm_fit <- ergm(ergm_formula, constraints = constraints) f <- paste0(\"Output/Singles/\", thresh, \"/ergmFit_\", sprintf(\"%03d\",i), \".RDS\") dir.create(dirname(f), showWarnings = FALSE, recursive = TRUE) saveRDS(bergm_fit, f)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "min and max age, mean and sd age, percentage of men and women [CODESPLIT] minAge = min(df_preQ$age) maxAge = max(df_preQ$age) meanAge = mean(df_preQ$age) sdAge = sd(df_preQ$age) females = length(which(df_preQ$gender == \"female\")) males = length(which(df_preQ$gender == \"male\")) other = length(which(df_preQ$gender == \"other\"))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Check normality with QQ plot and ShapiroWilk test Build the linear model [CODESPLIT] model <- lm(FW ~ condition, data = df_postQ)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Visualize correlations Insignificant correlations are leaved blank [CODESPLIT] corrplot(res_cor$r, method = 'number', type=\"upper\", order=\"hclust\", p.mat = res_cor$P, sig.level = 0.05, insig = \"blank\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot a correlation matrix with all the different measures Create a dataframe with all the different measures of above [CODESPLIT] df_corM <- data.frame(df_postQ$FW, df_postQ$DU, df_postQ$arousal, df_postQ$valence, df_postQ$control, df_postQ$responsibility, df_diff$Diff) colnames(df_corM) <- c('FW', 'DU', 'arousal', 'valence', 'control', 'responsibility', 'Diff')", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate posterior draws for regression lines [CODESPLIT] experiment_dat_virus <- dat[dat$material == material, ] print(experiment_dat_virus) experiment_dat_virus <- experiment_dat_virus %>% summarise(detection_limit = 10^first(detection_limit_log10_titer)) print(experiment_dat_virus) mat_dat <- dat[dat$material == material, ] if(material != \"Aerosols\") { scaling <- 1 ylab_expression <- expression(\"titer (TCID\"[50] * \"/mL media)\") max_x <- mat_dat %>% group_by(trial_unique_id, replicate) %>% filter(log10_titer == detection_limit_log10_titer) %>% select(time, trial_unique_id, replicate) %>% summarise(min_time = min(time)) %>% ungroup() %>% select(min_time) %>% max() } else { scaling <- 10 / 3 ## convert to tcid50/L air ylab_expression <- expression(\"titer (TCID\"[50] * \"/L air)\") max_x <- max(mat_dat$time) } print(max_x) plot_times <- dat %>% data_grid(time = seq_range(c(0, max_x), n = fineness)) print(material)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "draw n_lines random regression lines [CODESPLIT] func_samples <- tidy_draws %>% group_by(trial_unique_id, replicate) %>% sample_n(n_lines) %>% ungroup() print(func_samples)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "cross product decay_rates with x (time) values and calculate y (titer) values [CODESPLIT] to_plot <- func_samples %>% crossing(plot_times) to_plot <- to_plot %>% mutate(predicted_titer = scaling * 10^(intercept - decay_rate * time)) dat <- dat[dat$material == material, ] max_titer <- scaling * max(dat$titer)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "this function accepts a column containing hitnames, a target hitname to look for, and a ... starting and ending fixation index [CODESPLIT] is_hitname_in_range <- function(vec, hitname, fi_start, fi_end) { if (!\"fi_pairs\" %in% ls(envir = .GlobalEnv)) { fi_pairs <- get_fixationindex_pairs(df$FixationIndex)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "0 get sensing/location data for the specific user [CODESPLIT] sensing = dplyr::tbl(phonestudy, \"ps_activity\") %>% dplyr::filter(user_id %in% user) %>% dplyr::filter(!activityName %in% c(\"PHONESTUDY\", \"DNAPSOUND\", \"DNAPACCELEROMETER\", \"DNAPLIGHT\", \"DNAPPROXIMITY\", \"DNAPGYROSCOPE\")) %>% dplyr::mutate(timestamp = as.character(timestamp), created_at = as.character(created_at), updated_at = as.character(updated_at)) %>% data.frame() sensing$user_id = as.character(sensing$user_id)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Creates the variable 'distances', a 540 X 540 distance matrix between stimuli [CODESPLIT] coordinates = read.xlsx(\"540_coordinates.xlsx\", colNames = TRUE) coordinates = as.matrix(unname(coordinates)) coordinates = coordinates[,2:9] distances = as.matrix(dist(coordinates))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number of subjects in each group Create a group variable according to each cluster condition [CODESPLIT] if (size == 1){ N.size = N/K n.group = unlist(lapply(1:K, function(k) rep(k,N.size))) } if (size == 2){ if (K == 2){ N.1 = 0.10*N N.2 = N - N.1 n.group = c(rep(1,N.1),rep(2,N.2)) } if (K == 4){ N.1 = 0.10*N N.rest = N - N.1 N.size = N.rest/(K-1) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size),rep(4,N.size)) }} if (size == 3){ if (K == 2){ N.1 = 0.6*N N.2 = N - N.1 n.group = c(rep(1,N.1),rep(2,N.2)) } if (K == 4){ if (N == 20){ N.1 = 0.6*N N.rest = N - N.1 N.size = floor(N.rest/(K-1)) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size+1),rep(4,N.size+1)) } else{ N.1 = 0.6*N N.rest = N - N.1 N.size = N.rest/(K-1) n.group = c(rep(1,N.1),rep(2,N.size),rep(3,N.size),rep(4,N.size)) }}}", "target": 1, "target_options": ["no_match", "match"]}
{"input": "make a blank output table to hold the derive stats, and fill it up [CODESPLIT] out.tbl <- data.frame(array(NA, c(800, 5)));; # just make it big for now colnames(out.tbl) <- c(\"load.type\", \"cat.type\", \"trial.type\", \"stat.name\", \"stat.value\");; ctr <- 1;;", "target": 1, "target_options": ["no_match", "match"]}
{"input": "creating healthcare / containment variable [CODESPLIT] measures_new <- measures_new %>% mutate(measure_category = if_else(str_starts(measure,\"H\"), \"Healthcare\", \"Containment\"))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Number of subject in Group 0 [CODESPLIT] N0.subject = 2*N0.dyad", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exclude data from students who changed class or school during the school year. [CODESPLIT] quop2$exclusion[quop2$change>0] <- 1", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create date variable for birthday [CODESPLIT] quop2$s_birth_date <- as.Date(quop2$s_birth, format = \"%d.%m.%Y\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "set gavariables to NA if they have negative times where there are negative times recorded in the gqcvariables, use the corresponding gavariable value instead (hoping it is not negative) (!: ga also contains instruction pages) first and last item (one per page) between the instruction pages [CODESPLIT] js_start = c(2, 23, 37) js_end = c(21, 35, 49) for(i in 1:8){ for(j in 1:length(js_start)){ for(k in js_start[j]:js_end[j]){ eval(parse(text=paste0(\"quop_use$t\",i,\"_gqc\",k-j,\"<- ifelse(quop_use$t\",i,\"_gqc\",k-j,\"<=0,quop_use$t\",i,\"_ga\",k,\", quop_use$t\",i,\"_gqc\",k-j,\")\"))) } } }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Creat file paths for the raw data and the csv files that you get at the end [CODESPLIT] file_path = 'data_fullstudy1.txt' file_path2 = 'behavioral.csv' file_path3 = 'questionnaires.csv' file_path4 = 'boxes.csv'", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create loop for every participant, put the relevant data into the dataframe [CODESPLIT] con <- file(file_path, open = \"r\") on.exit(close(con)) lines <- readLines(con) subjects_data <- list() game_data <- list() boxes_data <- list() subj <- 0 for (line in lines) { if (grepl(\"Enter your Prolific ID\", line, fixed = TRUE)) { subj = subj + 1 metadata_beginning <- fromJSON(str_replace(line, \"\\\\}\\\\{\", \",\")) subj_data <- \"\"", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Recode NAs to 0 for response frequency tables [CODESPLIT] recodeNAto0 <- function(var) { case_when(is.na(var) ~ 0, T ~ var) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute response frequency table for recoded items [CODESPLIT] proptable <- function(var) { table(ds$country_name, var) %>% addmargins() %>% data.frame() %>% rename(country = 1, response = 2, prop = 3) %>% spread(response, prop) %>% mutate(`0` = `0`/Sum, `1` = `1`/Sum, `2` = `2`/Sum, `3` = `3`/Sum, `4` = `4`/Sum, `5` = `5`/Sum) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Make density ridge plot [CODESPLIT] ggplot(data = brmsmeans_antisci_cntr_post, aes(x = antisci_mean, y = fct_reorder(country_name, antisci_mean, .fun = mean), fill = factor(continent))) + stat_density_ridges(inherit.aes = T, calc_ecdf = F, quantile_lines = T, quantiles = 2, alpha = 0.7, scale = 3.5) + stat_summary(aes(label = paste0(sprintf(\"%0.2f\", round(..x.., 2)), \" [\", sprintf(\"%0.2f\", round(brmsmeans_hdi_l, 2)), \", \", sprintf(\"%0.2f\", round(brmsmeans_hdi_h, 2)), \"]\")), geom = \"text\", size = 3, hjust = 0, color = \"black\", position = position_nudge(x = 6.7 - brmsmeans_means)) + scale_fill_viridis_d(name = \"Continent\", option = \"inferno\", begin = 0.1, end = 0.9, direction = -1) + coord_cartesian(xlim = c(3.7, 6.3), clip = \"off\") + ggtitle(label = \"Posterior probability distributions of anti-science attitudes across countries\", subtitle = \"Annotations provide distribution means and 89% HDIs\") + xlab(\"Anti-science attitudes\") + ylab(\"\") + theme_minimal() + theme(plot.margin = unit(c(1, 10, 1, 1), \"lines\"), legend.position = \"bottom\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "extract the head and convert it to image [CODESPLIT] extract_pdf_top <- function(pdf_path) { page_img <- image_read_pdf(pdf_path, page = 1, density = 72) width <- image_info(page_img)$width height <- image_info(page_img)$height left <- 0 top <- 75 height <- height / 4 image_crop(page_img, geometry = sprintf(\"%fx%f+%f+%f\", width, height, left, top)) } dontcare <- first_page_paths %>% map(extract_pdf_top) %>% map2(png_paths, image_write)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Figures function to create data summaries in figures calculating means and cis for the plot [CODESPLIT] ci <- function(x) (sd(x)/sqrt(length(x))*qt(0.975,df=length(x)-1) ) #function to compute size of confidence intervals data_summary <- function(x) { m <- mean(x) ymin <- m-ci(x) ymax <- m+ci(x) return(c(y=m,ymin=ymin,ymax=ymax)) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "median thickness of new snow [CODESPLIT] medianThicknessNewSnow <- sapply(avgSP$sets, function(set) { median(sapply(set, function(sp) { sum(sp$layers$thickness[findPWL(sp, pwl_gtype = c(\"PP\", \"DF\"))]) })) }) lines(avgSP$meta$date, avgSP$meta$hs_median - medianThicknessNewSnow, lty = \"dashed\", lwd = 1) avgSP$avgs <- snowprofileSet(lapply(avgSP$avgs, function(avg) { avg$layers$percentage <- avg$layers$ppu_all avg })) plot(avgSP$avgs[avgSP$meta$date >= xdaterange[1] & avgSP$meta$date <= xdaterange[2]], ColParam = \"percentage\", add = TRUE, yaxis = FALSE, ylab = \"\") legend(as.Date(\"2018-09-23\"), 190, c(\"<HS>\", \"<NEW>\", \"PP\", \"DF\", \"SH\", \"DH\", \"FC\", \"FCxr\", \"RG\", \"MF\", \"MFcr\"), lty = c(\"solid\", \"dashed\", rep(NA, 9)), lwd = 2, col = c(rep(\"black\", 2), getColoursGrainType(c(\"PP\", \"DF\", \"SH\", \"DH\", \"FC\", \"FCxr\", \"RG\", \"MF\", \"MFcr\"))), pch = c(NA, NA, rep(15, 9)), pt.cex = 2.5, density = c(rep(0, 11)), border = \"transparent\", horiz = FALSE, bty = \"o\", box.lwd = 0, cex = 1.4) dev.off()", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Means and SDs of TOTAL LT TO THE SCREEN in the outcome phase (Okumura) Table 2 in the main manuscript: for mean.overall.data (analogously to the boxplots) [CODESPLIT] mean.overall.data %>% group_by(Condition) %>% summarise(mean=mean(LTScreenOut), sd=sd(LTScreenOut), na.rm = TRUE) %>% as.data.frame(.) %>% dplyr::mutate_if(is.numeric, round, 3)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "checking disctribution of the dependent variable [CODESPLIT] hist(overall.data$LTScreenOut) hist(overall.data$LTObjectOut) hist(overall.data$FirstLookDurationObjectOut)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "reduced model revealed a convergence warning, with high SDs for TrialRun and Concontext (random effects). model with simpler random effect structure revealed similar estimates suggesting that the warning can be ignored [CODESPLIT] red6_2 <- glmer(FirstLookDurationObjectOut ~ ConContext + Identity_change + Location_change + z.TrialRun + z.TrialCon + ObjectPosAct + (1 + z.TrialCon + ObjectPosAct | ID) , data=overall.data, family=Gamma(link=log), control=contr) summary(red6_2) summary(red6)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Test the difference between correlations of the same skills vs. different skills based on multiple imputation [CODESPLIT] names(Data_Table2a)[4]<-\"Bad_news\" m <- 'Persuasion ~~ v1*Persuasion + s1*Unreasonable + d1*Crisis + d2*Bad_news + d3*Presentation + d4*Mistake Unreasonable ~~ v2*Unreasonable + d5*Crisis + d6*Bad_news + d7*Presentation + d8*Mistake Crisis ~~ v3*Crisis + s2*Bad_news + d9*Presentation + d10*Mistake Bad_news ~~ v4*Bad_news + d11*Presentation + d12*Mistake Presentation ~~ v5*Presentation + s3*Mistake Mistake ~~ v6*Mistake", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compute boxplot characteristics [CODESPLIT] x <- boxplot(rep.int(yn, wn), plot = FALSE) top_vp <- viewport(layout = grid.layout(nrow = 2, ncol = 3, widths = unit(c(ylines, 1, 1), c(\"lines\", \"null\", \"lines\")), heights = unit(c(1, 1), c(\"lines\", \"null\"))), width = unit(1, \"npc\"), height = unit(1, \"npc\") - unit(2, \"lines\"), name = paste(\"node_boxplot\", nid, sep = \"\"), gp = gp) pushViewport(top_vp) grid.rect(gp = gpar(fill = bg, col = 0)) top <- viewport(layout.pos.col = 2, layout.pos.row = 1) pushViewport(top) if (is.null(mainlab)) { mainlab <- if (id) { function(id, nobs) sprintf(\"[%s] nWKL = %s\", # Node %s id, length(unique(VS_$wkl_uuid[fitted_node(obj$node, obj$data) == nid]))) # id, nobs", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate to and from percentages [CODESPLIT] from_id <- unique(AbbyLinks$from) for (i in 1:length(from_id)) { AbbyLinks$perc_from[AbbyLinks$from == from_id[i]] <- AbbyLinks$n[AbbyLinks$from == from_id[i]]/sum(AbbyLinks$n[AbbyLinks$from == from_id[i]]) } to_id <- unique(AbbyLinks$to) for (i in 1:length(to_id)) { AbbyLinks$perc_to[AbbyLinks$to == to_id[i]] <- AbbyLinks$n[AbbyLinks$to == to_id[i]]/sum(AbbyLinks$n[AbbyLinks$to == to_id[i]]) } rm(i, from_id, to_id) head(AbbyLinks, 10)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Generate Data Generate the outcome (y) and mediator (m) in the population with a correlation of b, variances of 1, and means of 0. [CODESPLIT] sigma <- rbind(c(1, b), c(b, 1)) mu <- c(0, 0) df <- as.data.frame(mvrnorm(n = n, mu = mu, Sigma = sigma)) names(df) <- c(\"y\", \"m\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a plot that illustrats: Accuracy, Kappa, AUROC and MCC [CODESPLIT] plot_3 <- methods_plot %>% filter(indices %in% c(\"Accuracy\", \"Kappa\", \"AUROC\" , \"MCC\")) ggplot(plot_3,aes(x=variable,y=value,fill=reorder(indices,value))) + geom_bar(stat = \"identity\",position = \"dodge\",color=\"black\") + xlab(\"\") + ylab(\"in %\") + ggtitle(\"\") + scale_fill_brewer(palette = \"Pastel1\",name=\"\") + theme_bw() + theme(legend.position=\"bottom\") + theme(legend.title=element_blank()) + theme(plot.title = element_text(hjust = 0.5)) + scale_y_continuous(breaks = c(0,12.5,25,37.5,50,62.5,75,87.5,100),limits = c(0,95)) + theme(axis.text.x = element_text(angle=30, hjust = 1, vjust = 1, size = 12)) + theme(axis.text.y = element_text( size = 12)) + theme(axis.title.y = element_text( size = 12)) + theme(legend.title = element_text( size = 12 ))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "round all numeric variables x: data frame digits: number of digits to round [CODESPLIT] numeric_columns <- sapply(x, class) == 'numeric' x[numeric_columns] <- round(x[numeric_columns], digits) x }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "M, SD, range of key sociodemographic variables (across countries) [CODESPLIT] summstats(ds, age, country_name) %>% print(n = nrow(.)) summstats(ds, sex_rec, country_name) %>% print(n = nrow(.)) summstats(ds, edu_9cat, country_name) %>% print(n = nrow(.)) summstats(ds, income_10cat, country_name) %>% print(n = nrow(.))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot 95% credible intervals [CODESPLIT] x<-plot(me, plot = FALSE)[[1]] + scale_color_grey() + scale_fill_grey() x+ylim(0,1)+ theme(panel.grid.major = element_line(colour=\"gray\"), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_blank(),panel.grid.major.x = element_blank())+xlab(\"\")+ylab(\"\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculating Bayes Factors for effects [CODESPLIT] bayes_factor(xfit1,xfit2) bayes_factor(xfit1,xfit3)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot of abstractlevel rating averages Scatterplot with overlaid box [CODESPLIT] eval.plot <- ggplot(abstract.plot, aes(y=eval, x=target.f, color=target.f)) + geom_jitter(aes(colour=target.f)) + geom_boxplot(outlier.shape = NA, alpha=0.5) eval.plot <- eval.plot + theme_bw() + scale_colour_manual(values=c(\"#9E2E2E\",\"royalblue4\")) + scale_x_discrete( labels = c(\"Conservatives\", \"Liberals\") ) + scale_y_continuous(limits = c(1, 7), expand = c(0,0)) + labs(y=\"Evaluative Rating\", x=\"\") + theme( text = element_text(size=16), panel.grid.major.x = element_blank(), legend.position = \"none\" ) explain.plot <- ggplot(abstract.plot, aes(y=explain, x=target.f, color=target.f)) + geom_jitter(aes(colour=target.f)) + geom_boxplot(outlier.shape = NA, alpha=0.5) explain.plot <- explain.plot + theme_bw() + scale_colour_manual(values=c(\"#9E2E2E\",\"royalblue4\")) + scale_x_discrete( labels = c(\"Conservatives\", \"Liberals\") ) + scale_y_continuous(limits = c(1, 7), expand = c(0,0)) + labs(y=\"Explanatory Rating\", x=\"\") + theme( text = element_text(size=16), panel.grid.major.x = element_blank(), legend.position = \"none\" ) grid.arrange(eval.plot,explain.plot, nrow = 1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot the raw data and the 95% HPDI from m.arithmetic [CODESPLIT] d.gg.math <- d.math.agg d.gg.math$mean <- NA d.gg.math$low_ci <- NA d.gg.math$high_ci <- NA", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixedeffects model for reproduction bias [CODESPLIT] mod_full <- lmer(Prct_Bias ~ ContextC*(GMSI_Gen_Z + GroupC*OrderC*Distort) + (ContextC | Subject), data=dat2) summary(mod_full) Anova(mod_full, type=3, test='Chisq') # Wald tests", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Simple slopes: Model for testing effect of Context at high GMSI [CODESPLIT] mod_1 <- lmer(Prct_Bias ~ ContextC*(GMSI_Gen_hi + GroupC*OrderC*Distort) + (ContextC | Subject), data=dat2) summary(mod_1) Anova(mod_1, type=3, test='Chisq')", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add informaiton on species + condition to each row of dplot [CODESPLIT] gramXspec <- as.factor(c(rep(\"Bird AD\", 3), rep(\"Bird NAD\", 3))) grammar <- as.factor(c(rep(\"AD\", 3), rep(\"NAD\", 3))) species <- as.factor(c(rep(\"Bird\", 6))) dplot $ Condition <- as.factor(grammar) ;; dplot$Species <- as.factor(species) ;; dplot $ gramXspec <- as.factor(gramXspec) dplot", "target": 1, "target_options": ["no_match", "match"]}
{"input": "exclude App Spirituality features (are accidentally still created in the data set as empty variables) [CODESPLIT] sensing = sensing %>% dplyr::select(!matches(\"Spirituality\"), -date)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Porportion of trials missed within each block, for each subj [CODESPLIT] miss.blk <- with(data, tapply(trialError, list(subj, data$block), function(x) sum(x!=\"FALSE\") / length(x)))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "plot fit model marginal_effects( FIT ) compute marginal means and average marginal effects [CODESPLIT] mu_happy = rowMeans(getCAT_fitted_mean(FIT , 'emotion' , 'happy' )) mu_angry = rowMeans(getCAT_fitted_mean(FIT , 'emotion' , 'angry' )) mu_neutral = rowMeans(getCAT_fitted_mean(FIT , 'emotion' , 'neutral' )) contrast_mu_angry_neutral = mu_angry-mu_neutral contrast_mu_happy_neutral = mu_happy-mu_neutral contrast_mu_angry_happy = mu_angry - mu_happy", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add column for centered logtransformed time series [CODESPLIT] habsos.ts$ctr <- habsos.ts$log10_CELLS - mean(habsos.ts$log10_CELLS) head( habsos.ts )", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Custom functions 1. Data management 1.1 Build mean indices from multiple variables [CODESPLIT] mean_index <- function (df, name, vars) { M1 <- dplyr::select(df, vars) M2 <- rowMeans(M1, na.rm = TRUE) M2 <- tibble::tibble(M2) colnames(M2) <- name df <- dplyr::bind_cols(df, M2) return(df) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "2. Stats 2.1 Calculate standard error (se_func) [CODESPLIT] se_func <- function(var) { sd <- sd(var) n <- length(var) se <- sd / sqrt(n) return(se) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate confidence interval (upper bound) [CODESPLIT] upper_ci_func <- function(var) { m <- mean(var) sd <- sd(var) n <- length(var) se <- sd / sqrt(n) lower_ci <- m + qt(1 - (0.05 / 2), n - 1) * se return(lower_ci) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "2.6 Bayes Factor paired ttest (func_ttest_paired_bf) [CODESPLIT] func_ttest_paired_bf <- function(pre = df_merge1_s3$paffect_pre, post = df_merge1_s3$paffect_post, dv = \"paffect\") {", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate pvalues from ttests [CODESPLIT] tee <- with(df, t.test(formula = dep_var ~ group_var, paired = FALSE, alternative = side)) return(tee) } tee <- with(df, t.test(x = var_pre, y = var_post, paired = TRUE, alternative = side)) } tee <- with(df, t.test(x = var_pre, y = var_post, paired = TRUE, alternative = side)) tee <- tee[[\"statistic\"]][[\"t\"]] effsize <- tee / sqrt(length(var_pre)) return(effsize) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "impute missing values with the median of the respective variable [CODESPLIT] phonedata <- impute(phonedata, target = \"Soci\", classes = list(numeric = imputeMedian(), integer = imputeMedian()))$data", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Print percents of stuff (used for displaying study demographics) [CODESPLIT] print.percents <- function(x) { t <- as.data.frame(sort(table(as.character(x)), decreasing = T)) t$Percent <- round(t$Freq/sum(t$Freq)*100) colnames(t) <- c(\"\", \"n\", \"%\") print(t) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effect model (RTs) 1000/RT as preregistered [CODESPLIT] LDTword_LME = lmer(-1000/RT ~ primec + (1|item) + (1+primec|subject), data = byTrial, contrasts = list(primec = cc1), control=lmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(LDTword_LME)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exploratory: Linear mixed effect model (RTs) with AuthorTest_Finnish [CODESPLIT] LDTword_LME_ARF = lmer(-1000/RT ~ primec*AR_Finnish + (1|item) + (1|subject), data = byTrial, contrasts = list(primec = cc1), control=lmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(LDTword_LME_ARF) plot(allEffects(LDTword_LME_ARF), x.var = \"AR_Finnish\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effect model (RTs) nonwords [CODESPLIT] LDTword_LMEnw = lmer(-1000/RT ~ primec + (1|item) + (1+primec|subject), data = byTrialnw, contrasts = list(primec = cc1), control=lmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(LDTword_LMEnw)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effect model (RTs) [CODESPLIT] Nword_LME = lmer(-1000/RT ~ primec + (1|item) + (1+primec|subject), data = nbyTrial, contrasts = list(primec = cc1), control=lmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(Nword_LME)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Linear mixed effect model (RTs) with ART_English [CODESPLIT] Nword_LME_AR = lmer(-1000/RT ~ primec*AR_English + (1|item) + (1+primec|subject), data = nbyTrial, contrasts = list(primec = cc1), control=lmerControl(optimizer=\"bobyqa\", optCtrl=list(maxfun=2e5))) summary(Nword_LME_AR) plot(allEffects(Nword_LME_AR), x.var = \"AR_English\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "exclude features with zero or nearzero variance [CODESPLIT] exclude_zero.var = nearZeroVar(data[, which(!colnames(data) %in% no.features)], freqCut = 95/5, uniqueCut = 10, saveMetrics = FALSE, allowParallel = TRUE) length(exclude_zero.var) exclude_zero.var = colnames(data[, which(!colnames(data) %in% no.features)])[exclude_zero.var] data = data %>% dplyr::select(-all_of(exclude_zero.var))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Add the posterior_samples values as a column in the dataframe. Repeat it as many times as there are cell in the factorial design, namely 2. factor estimates [CODESPLIT] post.data$b_GramGender2 <- rep(posterior_samples(model.study2)[['b_GramGender2']], 2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create dataset with only correct trials and only trials above 100 ms [CODESPLIT] data.3SD <- data[data$corr == 1, ] data.3SD <- data.3SD[data.3SD$min2000 == 1, ] data.3SD <- data.3SD[data.3SD$plus100 == 1, ]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Function for showing parameter estimats with pvalues for polr models [CODESPLIT] showCoefWithPValue <- function(model) { coef <- coef(summary(model)) pvalue <- pnorm(abs(coef[, \"t value\"]), lower.tail = FALSE) * 2 coef <- cbind(coef, \"pvalue\" = pvalue) return(round(coef, 4)) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Correlation between bulletin user type and avalanche awareness training [CODESPLIT] cor.test(as.numeric(PartBkgr$BullUseType), as.numeric(PartBkgr$BackgrAvTraining), method = \"spearman\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "This additional analysis test the influence of trialbytrial blink duration on response error by computing individual Bayes factors the Bayes factor is calculated by examining the residuals of a null model and how they correlate with blinkdurations the procedure use a 'default' prior for calculating bayes factors for correlations code here: [CODESPLIT] source(\"BF_correlations.R\") bf_null <- rep(NA, length(unique(d$Subject))) dur_slope <- rep(NA, length(unique(d$Subject))) for(i in 1:length(bf_null)){ d_i <- d[d$Subject==unique(d$Subject)[i] & d$cond1==1,] m0 <- lm(ResponseError~vel1, d_i) m1 <- lm(ResponseError~vel1 + cond1:bdur, d_i) dur_slope[i] <- coef(m1)[3] bf_null[i] <- 1/bf10JeffreysIntegrate(n=nrow(d_i), r=cor(residuals(m0), d_i$bdur)) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "median and range of BF supporting the null [CODESPLIT] round(median(bf_null[bf_null > 10^(1/2)]),digits=2) round(range(bf_null[bf_null > 10^(1/2)]),digits=2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "calculate number of days between symptoms [CODESPLIT] mutate( ill_where_n=ifelse(sym_temp==1, 1, 0), ill_where_n=cumsum(ill_where_n), ill_where_n=ifelse(sym_temp==1, ill_where_n, NA), prev_day_ill=lag(sym_temp, 1), day_lag=lag(day_date, 1),", "target": 1, "target_options": ["no_match", "match"]}
{"input": "recode illness vars as NA if no specific symptoms during episode [CODESPLIT] mutate_at(vars(ill_where_id, ill_where_n, ill_where_start, ill_where_end, ill_where_new), funs(ifelse(any_specif_sym==FALSE, NA,. ))) %>%", "target": 1, "target_options": ["no_match", "match"]}
{"input": "remove spaces, brackets, periods in column names [CODESPLIT] names(df) <- gsub(\"\\\\s+\", \"\", names(df)) names(df) <- gsub(\"\\\\(|\\\\)\", \"\", names(df)) names(df) <- gsub(\"\\\\.\", \"\", names(df)) cat(\" Ã°ÂÂÂ Column names fixed (no spaces, brackets, periods)\\n\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "initialization of the vectors with the names of the modueles/variables [CODESPLIT] names.modules.df.w <- names.var.modules <- NULL", "target": 1, "target_options": ["no_match", "match"]}
{"input": "list of the directories included in the datadir directory [CODESPLIT] my.dirlist <- list.dirs(path = datadir) if (length(my.dirlist) == 0) stop(\"There are no subdirectories in the directory that you selected.\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "creates a list with the names of the variables included for each module [CODESPLIT] names.var.modules <- vector(\"list\", num.waves) names(names.var.modules) <- paste(\"Wave\", waves) if(verbose) cat(\"Reading data from the selected modules within the waves. \\n\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "check which of the specified variables are actually present in the downloaded data [CODESPLIT] which.var.ok <- variables.in.modules[[i.mod]] %in% names(new.data) if (any(!which.var.ok)) warning( \"Some of the variables that you selected from module \", modules[i.mod], \" in wave \", waves[i.wave], \" are not included in the data set. \\n The variables that could not be imported were: \", variables.in.modules[[i.mod]][!which.var.ok], \". Please check these variables, the other variables were exported. \\n\" ) new.data <- select(new.data, variables.in.modules[[i.mod]][which.var.ok]) } names(new.data)[-1] <- paste(names(new.data)[-1], modules[i.mod], sep = \"...\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Trim outliers Perform winsorizing (replacefor each individual seperatelyvalues below 5th percentile and values above 95th percentile with 5th and 95th percentile values respectively) [CODESPLIT] for(i in min(all.df$person):max(all.df$person)){ all.df$zw.HR[all.df$person == i] <- Winsorize(all.df$z.HR[all.df$person==i]) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create a lag variable the data is lag within person and within days [CODESPLIT] lag.Y = function(data){ Y_lag = rep(0,nrow(data)) subjno.i = unique(data$subjno) for (i in subjno.i){ n.i = which(data$subjno==i) Y_lag[n.i] = shift(data$Y[n.i],1) } return(Y_lag) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Personality models Extract standardized parameters and N from personality models [CODESPLIT] Values_Analysis1_Model1_Pers <- Values_Personality[which(Values_Personality$analysis == \"analysis1\" & ((Values_Personality$study %in% c(\"S1\", \"S2\") & Values_Personality$DV == \"well_being\") | (Values_Personality$study == \"S3\" & Values_Personality$DV == \"affect_balance\")) & Values_Personality$output %in% c(\"standardized\", \"N\")), ] Values_Analysis1_Model2_Pers <- Values_Personality[which(Values_Personality$analysis == \"analysis2\" & ((Values_Personality$study %in% c(\"S1\", \"S2\") & Values_Personality$DV == \"well_being\") | (Values_Personality$study == \"S3\" & Values_Personality$DV == \"affect_balance\")) & Values_Personality$output %in% c(\"standardized\", \"N\")), ] Values_Analysis1_Model3_Pers <- Values_Personality[which(Values_Personality$analysis == \"analysis3\" & ((Values_Personality$study %in% c(\"S1\", \"S2\") & Values_Personality$DV == \"well_being\") | (Values_Personality$study == \"S3\" & Values_Personality$DV == \"affect_balance\")) & Values_Personality$output %in% c(\"standardized\", \"N\")), ]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Extract withinperson variances of interactions with close peers, family, and weak ties from unstandardized models [CODESPLIT] variances <- Values_Base$est[which(Values_Base$analysis == \"analysis3\" & ((Values_Base$study %in% c(\"S1\", \"S2\") & Values_Base$DV == \"well_being\") | (Values_Base$study == \"S3\" & Values_Base$DV == \"affect_balance\")) & Values_Base$output == \"unstandardized\" & Values_Base$paramHeader == \"Variances\" & Values_Base$param %in% c(\"PEERS\", \"FAMILY\", \"WEAK_TIES\") & Values_Base$BetweenWithin == \"Within\")] rows <- which(Values_Analysis1_Model3$paramHeader %in% c(\"S1|WB.ON\", \"S2|WB.ON\", \"S3|WB.ON\") & Values_Analysis1_Model3$param %in% c(\"PEERS\", \"FAMILY\", \"WEAK_TIES\")) # within-person effects variances <- Values_Personality$est[which(Values_Personality$analysis == \"analysis3\" & ((Values_Personality$study %in% c(\"S1\", \"S2\") & Values_Personality$DV == \"well_being\") | (Values_Personality$study == \"S3\" & Values_Personality$DV == \"affect_balance\")) & Values_Personality$output == \"unstandardized\" & Values_Personality$paramHeader == \"Variances\" & Values_Personality$param %in% c(\"PEERS\", \"FAMILY\", \"WEAK_TIES\") & Values_Personality$BetweenWithin == \"Within\")] rows <- which(Values_Analysis1_Model3_Pers$paramHeader %in% c(\"S1|WB.ON\", \"S2|WB.ON\", \"S3|WB.ON\") & Values_Analysis1_Model3_Pers$param %in% c(\"PEERS\", \"FAMILY\", \"WEAK_TIES\")) # within-person effects", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Model selection for four data sets First find bestfitting random effect model, then test for fixed effect predictors Process is backward by elimintating weakest terms sequentially, starting with full model, until only significant effects remain Nested model comparisons (likelihood ratio tests using anova command) are used to select best fitting models _ Pair Experiment 1 Random effects: selecting empty model [CODESPLIT] pair1_rand_full <- lmer(distance ~ (1 | group) + (1 + session0 | pair), data = pair_data1) # full random model;; fails to converge pair1_rand_full <- lmer(distance ~ (1 | group) + (1 + session0 | pair), data = pair_data1, control = lmerControl(optimizer = \"bobyqa\")) # full random model;; uses bobyqa optimizer pair1_rand2 <- lmer(distance ~ (1 + session0 | pair), data = pair_data1) # drops group ran.intercept;; fails to converge pair1_rand2 <- lmer(distance ~ (1 + session0 | pair), data = pair_data1, control = lmerControl(optimizer = \"bobyqa\")) # drops group ran.intercept;; uses bobyqa optimizer;; BEST pair1_rand3 <- lmer(distance ~ (1 | pair), data = pair_data1) # drops random slope", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Pearson's correlation calculation [CODESPLIT] calc_r <- function(model_vector, data_vector, S_0, D_0, M_0, E_0, filename) { sst_mean <- mean(data_vector) sst_vector <- (data_vector - sst_mean) ^ 2 sst <- sum(sst_vector) ssr_vector <- (data_vector - model_vector) ^ 2 ssr <- sum(ssr_vector) r_sq <- 1 - (ssr / sst) cat(r_sq, file = paste(format(Sys.time(),\"%Y_%m_%d_%H_%M\"), filename, \"r_sq\", \"S\", S_0, \"D\", D_0, \"M\", M_0, \"E\", E_0, \".txt\", sep = \"_\")) return(r_sq) } r_sq <- calc_r(CO2_flux_ratios_hat_median, data_vector, S_0, D_0, M_0, E_0, filename)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Data wrangling recode values for missing data (9, 1) in whole dataset as NA [CODESPLIT] data <- data %>% mutate_all(~na_if(., -9)) data <- data %>% mutate_all(~na_if(., -1))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "age recode values of 99 as missing [CODESPLIT] data$indager.w2[data$indager.w2 == 99] <- NA data$age <- data$indager.w2", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit models with trimmed weights Weighted regression model for SciPop Score (Goertz approach) [CODESPLIT] m3.scipopgoertz.trim.svyglm.fit <- svyglm(scipopgoertz ~ age + gender + education.comp + education.uni + sciprox.score + urbanity.log + languageregion.ger + languageregion.ita + polorientation + religiosity + interestscience + sciliteracy + trustscience + trustscientists, design = bar.design.scipopgoertz.trim, family = gaussian, na.action = na.omit)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "for each row assign fixation probability (1 or 0) to each picture based on the activation !! need a better decision criteria when the activations are the same;; !! right now it assigns '1' to both [CODESPLIT] dt.act$max.act = apply(dt.act[,4:7], 1, max) dt.act = subset(dt.act, max.act != 0)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "4.4.4) Create data where u(x) is at sample means to get residuals based on rest of models to act as yobs Recall: columns 1 & 2 have y and u(x) in obs.data [CODESPLIT] data.xufixed =data.obs data.xufixed[,2]=mean(data.obs[,2]) #Note, the 1st predictor, 2nd columns, is always the one hypothesized to be u-shaped", "target": 1, "target_options": ["no_match", "match"]}
{"input": "6 RUN TWO LINE REGRESSIONS 6.1 First an interrupted regression at the midpoint of the flat region [CODESPLIT] rmid=reg2(f,xc=median(xflat),graph=0, axislabels=axislabels)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Pairwise comparison of optimism scores for each cue and table Wilcoxon signedrank test [CODESPLIT] cue_diff_wilcox_ts = wilcox.paired.multcomp( opt_score ~ cue | id, subset(jbt_all, jbt_all$cjb_test == \"ts\"), p.method = \"holm\") cue_diff_wilcox_tunnel = wilcox.paired.multcomp( opt_score ~ cue | id, subset(jbt_all, jbt_all$cjb_test == \"tunnel\"), p.method = \"holm\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "round the numeric values (p that is not < 0.001) [CODESPLIT] cue_diff_wilcox_ts$p.value[numeric_values] <- round(as.numeric(cue_diff_wilcox_ts$p.value[numeric_values]), 3) names(cue_diff_wilcox_ts) <- c(\"Cues compared\",\"p-value\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "new tibble with IDs and emails [CODESPLIT] wos_emails <- tibble( \"responseID\" = wos$response_ID, \"email\" = wos$email ) aaas_emails <- tibble( \"responseID\" = aaas$response_ID, \"email\" = aaas$email )", "target": 1, "target_options": ["no_match", "match"]}
{"input": "drop rows with empty email fields [CODESPLIT] wos_emails <- drop_na( wos_emails, email ) aaas_emails <- drop_na( aaas_emails, email )", "target": 1, "target_options": ["no_match", "match"]}
{"input": "getmode() Get mode of a variable Input: numeric vector Output: single number [CODESPLIT] getmode <- function(v) { uniqv <- unique(v) uniqv[which.max(tabulate(match(v, uniqv)))] }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "For studyProgramme remove countries with only one programme [CODESPLIT] if(varname == \"studyProgram\" & length(lev) == 1){ return(NA) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Model for Frequency DV [CODESPLIT] mod1.1 <- lmer(frequency ~ ruleType + (1|sceneType) + (1|Participant), contrasts = my.contrasts, data = PD3a.Long) icc(mod1.1) summary(mod1.1) emmeans(mod1.1,list(pairwise~ruleType),adjust=\"satterthwaite\") confint(mod1.1)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Define functions Means, SDs, and two sample ttests comparing Swiss census data and the panel sample [CODESPLIT] census.msd <- function(v) { v <- reformulate(v) cbind( as.data.frame(svyby(v, ~t, dslong_dsgn, svymean, na.rm = T))[-c(1,3)], # M (exclude SE) sqrt(as.data.frame(svyby(v, ~t, dslong_dsgn, svyvar, na.rm = T)[[2]]))) %>% # SD (with a workaround) set_colnames(c(\"M\", \"SD\")) %>% set_rownames(c(\"2019 (census)\", \"2019 (panel)\")) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "RQ1: Examine predictors of change in SciPop Scores and subscale scores .. Visual inspection with Sankey plot Visualize withinsubject change in SciPop Score in Sankey plot [CODESPLIT] dslongpnl %>% group_by(t, scipopgoertz) %>% summarise(scipopgoertz_freq = n()) %>% merge(dslongpnl, by = c(\"scipopgoertz\", \"t\")) %>% transform(scipopgoertz_lvls = factor(scipopgoertz, levels = c(\"5\", \"4.5\", \"4\", \"3.5\", \"3\", \"2.5\", \"2\", \"1.5\", \"1\", \"NA\"))) %>% replace_na(list(scipopgoertz_lvls = \"NA\")) %>% ggplot(aes(x = t, stratum = scipopgoertz_lvls, alluvium = id, y = scipopgoertz_freq, fill = scipopgoertz_lvls, label = scipopgoertz_lvls)) + scale_x_discrete(expand = c(.1, .1)) + scale_fill_viridis_d() + geom_flow(width = .1) + geom_stratum(alpha = .5, width = .1) + geom_text(stat = \"stratum\", size = 4) + ggtitle(\"Sankey plot of within-subject change in SciPop Score\") + theme_minimal() + theme(legend.position = \"none\", panel.grid = element_blank(), axis.text.y = element_blank(), axis.ticks.y = element_blank())", "target": 1, "target_options": ["no_match", "match"]}
{"input": "H1 analyses .. Descriptive analyses Plot means, SDs, skewness, kurtosis of alternative SciPop Scores [CODESPLIT] meantbl <- as.data.frame(matrix(0, nrow = 2)) groups <- quos(scipopgoertz, scipopbollenm, scipopbollencfa, scipopsartori75, scipopsartoricat) for (j in seq_along(groups)) { meantbl <- cbind(meantbl, group_by(dslongpnl, t) %>% summarise(mean = mean(!!groups[[j]], na.rm = T), sd = sd(!!groups[[j]], na.rm = T), skewness = skewness(!!groups[[j]], na.rm = T), kurtosis = kurtosis(!!groups[[j]], na.rm = T))) names(meantbl)[names(meantbl) == \"mean\"] <- paste0(quo_text(groups[[j]]), \" (M)\") names(meantbl)[names(meantbl) == \"sd\"] <- paste0(quo_text(groups[[j]]), \" (SD)\") names(meantbl)[names(meantbl) == \"skewness\"] <- paste0(quo_text(groups[[j]]), \" (Skewness)\") names(meantbl)[names(meantbl) == \"kurtosis\"] <- paste0(quo_text(groups[[j]]), \" (Kurtosis)\") meantbl <- select(meantbl, contains(\"scipop\")) %>% set_rownames(c(\"2019\", \"2020\")) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "save means and compute their higher order terms [CODESPLIT] dfs$x.mean <- mean( dfs$x, na.rm = TRUE ) dfs$y.mean <- mean( dfs$y, na.rm = TRUE ) dfs$x2.mean <- dfs$x.mean^2 dfs$xy.mean <- dfs$x.mean*dfs$y.mean dfs$y2.mean <- dfs$y.mean^2 dfs }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "add column indicating if individual gave birth during testing (this will help with age classification) [CODESPLIT] birthsanc <- read_excel(\"motherssanctuary.xlsx\", col_types=c(\"text\", \"date\")) bigdatasanctuary$Individual<-as.factor(bigdatasanctuary$Individual) birthsanc$Mother <- as.factor(birthsanc$Mother) bigdatasanctuary$mother <- ifelse(bigdatasanctuary$Individual %in% birthsanc$Mother, \"yes\", \"no\") bigdatasanctuary$birth <- birthsanc$Date[match(bigdatasanctuary$Individual, birthsanc$Mother)] bigdatasanctuary$birth <- as.Date(bigdatasanctuary$birth, format=\"YY-mm-dd\") bigdatasanctuary$testbeforebirth <- ifelse(!is.na(bigdatasanctuary$birth) & bigdatasanctuary$birth > bigdatasanctuary$TestDate, \"yes\", \"no\") bigdatasanctuary$AgeAtTesting <- as.numeric(bigdatasanctuary$AgeAtTesting)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Step 1 compare the rising ridge asymmetric congruence model (RRCA) and the full thirdorder model (cubic) [CODESPLIT] rrca_comp <- compare2(rrca_myrsa, \"RRCA\", \"cubic\") rrca_comp", "target": 1, "target_options": ["no_match", "match"]}
{"input": "How much variance in the outcome can be explained by the full thirdorder polynomial model/the rising ridge leveldependent congruence model? [CODESPLIT] getPar(rrcl_myrsa, model=\"cubic\", type=\"R2\") getPar(rrcl_myrsa, model=\"RRCL\", type=\"R2\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get subsequent fixation index (+1) and retrieve row number using fi_pars and assign [CODESPLIT] scope_start[i] <- fi_pairs$fistart[scope_start_fis[i] + 1] } return(scope_start) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Exclude participants who completed less than 10 surveys [CODESPLIT] dat <- dat[-which(dat$N < 10), ] dim(dat) # 36074 assessments length(unique(dat$id)) # 1177 participants (1109 participants excluded)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "define function to compute colors and point sizes that yield 3d effect [CODESPLIT] points3d <- function(data, xname, yname, zname, cex.limit.close = 1.7, cex.limit.far = 1, col.limit.close = 0.1, col.limit.far = 0.8, adapted_eye = list(x=-12, y=-16, z=2) ){", "target": 1, "target_options": ["no_match", "match"]}
{"input": "for each row of the data, compute euclidean distance between the person's 3dim point and the plane [CODESPLIT] distance <- abs( (data[,xname] - a$x) * n$x + (data[,yname] - a$y) * n$y + (data[,zname] - a$z) * n$z ) / sqrt(n$x^2 + n$y^2 + n$z^2)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Calculate delta change in testosterone pivot data sets for delta values [CODESPLIT] hormones_wide <-hormones %>% pivot_wider(id_cols = c(\"id\", \"sex\"), names_from = condition, values_from = t_conc_corr) %>% mutate(delta_t = back_home - baseline) delta_t <- left_join(hormones, hormones_wide) delta_t <- delta_t %>% filter(condition == \"baseline\")%>% filter(!is.na(delta_t))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Plot sex and time point Boxplot for sex differences [CODESPLIT] t_condition <- hormones %>% ggplot(aes(x = condition, y = log(t_conc_corr), fill = sex)) + geom_boxplot(outlier.shape = NA, width = 0.7) +", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Network Estimation Ising Model (Binary Data) can be estimated via estimateNetwork variables automatically binarized at median estimate regularized logistic nodewise regression network define where to binarize variables eLASSO (LASSO with EBIC model selection) listwise deletion of missing values (pairwise not possible for regressions) [CODESPLIT] Ising_net <- estimateNetwork(data, default = \"IsingFit\", missing = \"listwise\", rule = \"OR\")", "target": 1, "target_options": ["no_match", "match"]}
{"input": "compare GGM and Ising model correlating weights matrices [CODESPLIT] cor.test(Wmat_GGM[upper.tri(Wmat_GGM)], Wmat_Ising[upper.tri(Wmat_Ising)]) #.80", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a storage container (i.e. a empty lists) for all hit_names ... ... that track looking times over all trials (e.g., looking_times$left) [CODESPLIT] looking_times <- setNames(vector(\"list\", length(hit_names)), hit_names)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a storage container for all looking frequencies (i.e., counting the number of looks within an AOI) [CODESPLIT] looking_frequencies <- setNames(vector(\"list\", length(hit_names)), hit_names)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "init storage containers for looking frequencies [CODESPLIT] current_trial_total_looks <- setNames(vector(\"list\", length(hit_names)), hit_names) for (hn in hit_names) {", "target": 1, "target_options": ["no_match", "match"]}
{"input": "get first and last FixationIndex (remove NAs), which define the boundaries of a single trial [CODESPLIT] min_FixationIndex <- min(inter_trial_FixationIndexes, na.rm = TRUE) max_FixationIndex <- max(inter_trial_FixationIndexes, na.rm = TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "space between axis label and tick mark labels [CODESPLIT] my_settings$layout.widths$ylab.axis.padding <- 0.2 my_settings$layout.heights$axis.xlab.padding <- 0.2 my_settings$box.rectangle$col = 1 my_settings$box.umbrella$col = 1 my_settings$box.dot$col = 1 my_settings$plot.symbol$col = 1", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Overlay the basic grid with colourful grid corresponding to the limits of sensitive period rectangles. These colourful lines should start in the points indicating the change values. [CODESPLIT] cols<-colscale[sapply(result$p1,function(x){which.min(abs(x-scaleseq))})] cols<-colscale[sapply(result$p1.rand[,exrun],function(x){which.min(abs(x-scaleseq))})]", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Skew effect [CODESPLIT] neg <- apply(chains[,c(1,3)], 1, sum) pos <- apply(chains[,c(1,2)], 1, sum) quantile(neg-pos, prob=c(.025, .975)) neg <- apply(chains[,c(1,3)], 1, sum) pos <- apply(chains[,c(1,2)], 1, sum) quantile(pos-neg, prob=c(.025, .975))", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Fit hierarchical Bayesian model [CODESPLIT] fit.HBM <- stan(file='CAM_full_1.stan', data=c('Nobs','Nind','stim','bias','RM','AS','B1','id','mid','stim_sizes','Nstim'), chains=4, iter=2500, cores=no_cores, control=list(adapt_delta=.90, max_treedepth=15)) fit.HBM sum(summary(fit.HBM)$summary[,'Rhat'] > 1.01) pars <- extract(fit.HBM)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Compute probabilities using MNL model [CODESPLIT] P[['choice']] = apollo_mnl(mnl_settings1, functionality) P[[\"indic_cost_tap\"]] = apollo_mnl(mnl_settings2, functionality) P[[\"indic_na1\"]] = apollo_mnl(mnl_settings3, functionality) P[[\"indic_na2\"]] = apollo_mnl(mnl_settings4, functionality) P[[\"indic_na3\"]] = apollo_mnl(mnl_settings5, functionality) P[[\"indic_cost_bottle\"]] = apollo_ol(ol_settings5, functionality) P[[\"indic_bill\"]]= apollo_normalDensity(normalDensity_settings1,functionality) P[[\"indic_qual\"]] = apollo_ol(ol_settingsB1, functionality) P[[\"indic_qual_f\"]] = apollo_ol(ol_settingsB3, functionality) P = apollo_combineModels(P, apollo_inputs, functionality)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "d) Ftest (Stephan Gries 2013, p. 218) [only for normally distributed variables!] use var.test() Output interpretation: 1) Is p > .05?;; 2) Does the CI include 1? [CODESPLIT] F.lextale <- var.test(pp.explicit.incidental$lextale~pp.explicit.incidental$learningtype);; F.lextale", "target": 1, "target_options": ["no_match", "match"]}
{"input": "c) Wilcoxon ranksum test ( MannWhitney U test) [Nonparametric alternative if normality is violated, or if you have ordinal data] Continuous variables: ratioscaled lextale [CODESPLIT] wilcox.lextale <- wilcox.test(lextale~learningtype, data=pp.explicit.incidental, paired=FALSE);; wilcox.lextale", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Selfrated variables: ordinal proficiency_overall [CODESPLIT] wilcox.proficiency_overall <- wilcox.test(proficiency_overall~learningtype, data=pp.explicit.incidental, paired=FALSE);; wilcox.proficiency_overall", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Effect size (approximate) for Wilcoxon ranksum test Write the function (from Field, Miles & Fiels 2012, p.665) [CODESPLIT] rWilcox <- function(wilcoxModel, N){ z <- qnorm(wilcoxModel$p.value/2) r <- z/sqrt(N) cat(wilcoxModel$data.name, \"Effect size, r = \", r) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create new variable for the IRV intraindividual response variability [CODESPLIT] data$irv <- irv( dplyr::select( data, belonging, control, meaningful_existence, self_esteem ), na.rm = TRUE, split = FALSE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "display the correlations as a histogram and heatmap [CODESPLIT] cor_matrix_half <- cor_matrix[upper.tri(cor_matrix)] mean(cor_matrix_half) sd(cor_matrix_half) hist(as.vector(cor_matrix_half), breaks=24, cex.axis=2) # Note: Novich et al. suppressed correlations of r<.4 in their visualisation heatmap(x = cor_matrix, symm = TRUE)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "Create stringent dataset Exclude People with Strange Response Patterns These are identified as subclusters in the inclusive dendogram that group together [CODESPLIT] inclusive_data$weird_responses = inclusive_data$body_postures_shape + inclusive_data$punctuation_shape + inclusive_data$letter_shape + inclusive_data$number_shape + inclusive_data$people_name_shape + inclusive_data$english_word_shape + inclusive_data$foreign_word_shape + inclusive_data$tastes_taste + inclusive_data$smells_smell + inclusive_data$noises_noise + inclusive_data$music_music + inclusive_data$colour_colour + inclusive_data$shapes_shape + inclusive_data$smells_taste + inclusive_data$tastes_smell + inclusive_data$voices_noise + inclusive_data$voices_music + inclusive_data$noises_music + inclusive_data$music_noise hist(inclusive_data$weird_responses) table(inclusive_data$weird_responses)", "target": 1, "target_options": ["no_match", "match"]}
{"input": "gives the prevalence of each cluster of data [CODESPLIT] colMeans(short_data) N_types <- matrix(apply(short_data[,1:i], 1, sum, na.rm=TRUE)) stringent_N_clusters <- cbind(stringent_N_clusters,N_types) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a continuoius time metric (seconds since midnight of day0) [CODESPLIT] dat$sec_midnight.day0 <- dat$start.secmidnight + dat$day * 86400", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create a variable indicating, if the current observation is the first observation of the day (\"morning\") this variable is 1, if the current observation was obtained on a different day than the previous observation [CODESPLIT] dat[!is.na(dat$lagday) & dat$day!=dat$lagday, \"morning\"] <- 1 dat[!is.na(dat$lagday) & dat$day==dat$lagday, \"morning\"] <- 0", "target": 1, "target_options": ["no_match", "match"]}
{"input": "create continuous time variable in hours and seconds [CODESPLIT] data$time_to_hours = lubridate::hour(data$timestamp.corrected) + lubridate::minute(data$timestamp.corrected)/60 + lubridate::second(data$timestamp.corrected)/3600 data$time_to_sec = data$time_to_hours*60*60 return(data) }", "target": 1, "target_options": ["no_match", "match"]}
{"input": "set font size for facet labels [CODESPLIT] strip.text.x = element_text(size = font_size_facets_x), strip.text.y = element_text(size = font_size_facets_y), strip.text.x = element_text(size = font_size_facets_x), strip.text.y = element_text(size = font_size_facets_y),", "target": 1, "target_options": ["no_match", "match"]}