gesterestr = read.csv(file="frequencies_Geste_full.csv", sep = ";", header = TRUE, row.names=1, quote = '\"')
gesterestr = as.matrix(gesterestr)
monCorpus = gesterestr
Some general description
#Total tokens
sum(gesterestr)
## [1] 1104296
#Total forms
nrow(gesterestr)
## [1] 52202
#Most frequent
sum(gesterestr[1,])
## [1] 34115
#Verif. Zipf
sum(gesterestr[10,])
## [1] 13712
#Number of hapaxes
nrow(gesterestr[rowSums(gesterestr) == 1,])
## [1] 25811
#%age hapax
nrow(gesterestr[rowSums(gesterestr) == 1,]) / nrow(gesterestr) * 100
## [1] 49.44447
monCorpus = gesterestr
#Graphe de dispersion
plot(colSums(monCorpus), ylab = "Number of words", main="Scatter plot", sub = nomCorpus)
#Histogramme
hist(colSums(monCorpus), main = "Number of words", sub=nomCorpus, xlab = "Number of words", ylab = "Frequency")
#Histogramme un peu plus configuré
#hist(colSums(monCorpus), breaks=seq(1000,30000,1000), main = "Nombre de mots par texte", sub=nomCorpus, xlab = "Nombre de mots", ylab = "Fréquence")
Description:
#Summary
summary(colSums(monCorpus))
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 387 7016 11490 22086 28366 217942
#SD
sd(colSums(monCorpus))
## [1] 32507.24
#Variance
var(colSums(monCorpus))
## [1] 1056720966
Power-law distribution.
#Moyenne géométrique
exp(mean(log(colSums(monCorpus))))
## [1] 12016.18
Boxplot,
boxplot(colSums(monCorpus), main = "Number of words per text", ylab="Number of words", sub=nomCorpus)
Texts ranked
colSums(monCorpus)[order(colSums(monCorpus), decreasing = TRUE)]
## lorr_1387pm13_pic_1365ca_BaudSebC
## 217942
## pic_1213pm13_pic_1190pm10_AliscW
## 71752
## pic_1213pm13_pic_1190pm10_MonRaincB
## 53557
## pic_1275pm25_pic_1210pm10_Aiol2N
## 49242
## art_1295_picmérid_1180ca_MonGuill1C2
## 48970
## nil_1450pm10_pic_1300ca_EnfGarB.
## 42516
## pic_1275pm25_pic_1160caAiol1N
## 41408
## agn_1225pm25_agn_1170ca_HornP.C
## 39489
## nil_1250pm50_nil_1230ca_GuiBourgG
## 35800
## bourg_1270ca_nil_1210pm10_AimeriD
## 33433
## Paris_1290pm10_flandr_1275_AdenBuevH
## 33077
## agn_1137pm13_Nord.Ouest_1100caRolS
## 29123
## bourg_1270ca_nil_1213pm13_MortAymC
## 28997
## agn_1250pm10_agn_1150pm16_ChGuillM
## 26472
## lorrsept_1275pm25_nil_1200ca_AmAmD
## 25291
## pic_1275pm25_pic_1190pm10ElieB.
## 24052
## bourg_1325pm25_Sud.Est_1190pm10_FloovG
## 21392
## frc_1262pm13_nil_1150pm16_CourLouisLe
## 19632
## Nord_1275pm25_Nord.Est_1190pm10RCambr2M
## 17913
## lorr_1290pm10_picmérid_1225caOrsonP
## 17084
## agn_1275pm25_nil_nil_Otin_B
## 13607
## Nord.Est_1262pm13_Nord.Est_1190pm10_PriseOrabR1
## 13561
## agn_1290pm10_agn_1250pm10_DestrRomeF2
## 13269
## StBrieuc_1317_nil_nil_Otin_A
## 12140
## Paris_1335ca_nil_1150pm17CharroiSch_B1.
## 11685
## Paris_1335ca_nil_1150pm17CharroiSch_B2.
## 11295
## pic_1225pm25_Nord.Est_1190pm10RCambr1M
## 11053
## lorrmérid_1275pm25_nil_1150pm20CharroiSch_D.
## 10832
## art_1295_nil_1150pm17CharroiSch_C.
## 10798
## frc_1263pm13_nil_1150pm17CharroiSch_A2.
## 10688
## frc_1283pm17_nil_1150pm17CharroiSch_A4.
## 10542
## frc_1263pm13_nil_1150pm17CharroiSch_A1.
## 9574
## lorr_1325pm25_nil_1150pm17CharroiSch_A3.
## 8743
## agn_1290pm10_agn_1175pm25_PelCharlB
## 7588
## bourg_1270ca_champmérid_1210pm10GirVianeE
## 7403
## nil_1300ca_norm_1200caAyeB
## 7176
## Meuse_1262pm13_lorr_1200caPriseCordD
## 7117
## Est_1300ca_pic_1213pm13FlorenceW
## 6983
## pic_1213pm13_picmérid_1150pm16_MonGuill1C1
## 6927
## agn_1335ca_agn_1190pm10_AmAmOctF
## 6888
## agn_1250pm10_agn_1170ca_HornP.O
## 6592
## bourg_1270ca_frc_1210pm10GuibAndrM
## 5725
## pic_1290pm10_pic_1275pm25FlorOctOctV
## 4589
## agn_1213pm13_frc_1125pm25GormB
## 3798
## agn_1200pm20_agn_1180pm10_Asprem.P4
## 2369
## agn_1200pm20_nil_nil_Otin_M
## 2085
## StBrieuc_1317_nil_1190ca_Fier.V
## 1682
## agn_1250pm50_nil_1250pm50_MacaireAl2B
## 1179
## nil_1250pm50_nil_1150pm17CharroiSch_fragm.
## 879
## agn_1250pm16_agn_1180pm10_Asprem.C
## 387
plot(rowSums(monCorpus), ylab = "Number of occurrences", main="Scatter plot", sub = nomCorpus)
hist(rowSums(monCorpus), breaks=10000, main="Word frequencies", sub=nomCorpus, xlab = "Number of occurrences", xlim=c(1,200))
#Summary
summary(rowSums(monCorpus))
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.00 1.00 2.00 21.15 4.00 34115.00
#SD
sd(rowSums(monCorpus))
## [1] 364.8223
#Variance
var(rowSums(monCorpus))
## [1] 133095.3
Power-law type distribution. Frequencies on a logarithmic scale:
tableFreq = table(rowSums(monCorpus))
freqCounts = cbind(as.numeric(labels(tableFreq)[[1]]), as.vector(tableFreq))
plot(freqCounts[,1], freqCounts[,2], xlab = "word freq.", ylab = "nb. of forms with this freq.")
#logarithmic scale
plot(freqCounts[,1], freqCounts[,2], xlab = "word freq.", ylab = "nb. of forms with this freq.", log="xy")
Geometric mean:
exp(mean(log(rowSums(monCorpus))))
## [1] 2.565113
Some analysis on MFW:
#Total Frequency rank 1
sum(gesterestr[1,])
## [1] 34115
#Total Frequency rank 600
sum(gesterestr[600,])
## [1] 201
#Total Frequency rank 1000
sum(gesterestr[1000,])
## [1] 114
#Total Frequency rank 1000
sum(gesterestr[1200,])
## [1] 94
#Total Frequency rank 2000
sum(gesterestr[2000,])
## [1] 52
#Total Frequency rank 2500
sum(gesterestr[2500,])
## [1] 40
#Total Frequency rank 3000
sum(gesterestr[3000,])
## [1] 32
#Means by text: 600
sum(gesterestr[600,]) / ncol(gesterestr)
## [1] 4.02
#Means by text: 1000
sum(gesterestr[1000,]) / ncol(gesterestr)
## [1] 2.28
#Means by text: 2000
sum(gesterestr[2000,]) / ncol(gesterestr)
## [1] 1.04
monCorpus2 = read.csv(file="frequencies_Geste_ProperNamesRemoved_for1-4000.csv", sep = ";", header = TRUE, row.names=1, quote = '\"')
monCorpus2 = as.matrix(monCorpus2)
Relative frequencies for the 600 MFW
#Faire varier le chiffre en dessous pour tenter d'autres sélections
monCorpusSelect = monCorpus2[1:600,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library('FactoMineR')
maBase = t(monCorpusSelect)
monACP = PCA(maBase)
monACP$eig
## eigenvalue percentage of variance
## comp 1 72.9928920 12.16548200
## comp 2 45.4577306 7.57628843
## comp 3 34.4948518 5.74914197
## comp 4 30.0493978 5.00823296
## comp 5 28.4858245 4.74763742
## comp 6 24.3679325 4.06132208
## comp 7 23.1597676 3.85996127
## comp 8 21.1972684 3.53287807
## comp 9 20.5741798 3.42902996
## comp 10 17.6059818 2.93433030
## comp 11 16.1896124 2.69826873
## comp 12 15.1803819 2.53006365
## comp 13 14.1698426 2.36164044
## comp 14 13.0110835 2.16851392
## comp 15 12.2731525 2.04552542
## comp 16 11.9545830 1.99243050
## comp 17 11.4615260 1.91025434
## comp 18 11.1876788 1.86461313
## comp 19 10.7945443 1.79909071
## comp 20 10.6250330 1.77083884
## comp 21 10.3052821 1.71754702
## comp 22 9.8947067 1.64911778
## comp 23 9.2025950 1.53376584
## comp 24 8.9688082 1.49480137
## comp 25 8.7942087 1.46570146
## comp 26 8.1482268 1.35803781
## comp 27 7.8933084 1.31555139
## comp 28 6.9736568 1.16227614
## comp 29 6.7999545 1.13332575
## comp 30 6.6952829 1.11588049
## comp 31 6.2335393 1.03892321
## comp 32 6.0701630 1.01169384
## comp 33 5.8794697 0.97991162
## comp 34 5.5426312 0.92377187
## comp 35 5.4544902 0.90908170
## comp 36 5.0988085 0.84980142
## comp 37 4.6157259 0.76928765
## comp 38 4.5431727 0.75719545
## comp 39 4.1871575 0.69785959
## comp 40 3.8282652 0.63804420
## comp 41 3.6753772 0.61256286
## comp 42 3.3125479 0.55209131
## comp 43 3.2549854 0.54249756
## comp 44 2.4111989 0.40186648
## comp 45 2.0952063 0.34920105
## comp 46 1.5706623 0.26177705
## comp 47 1.5183192 0.25305320
## comp 48 1.2404399 0.20673998
## comp 49 0.5585449 0.09309081
## cumulative percentage of variance
## comp 1 12.16548
## comp 2 19.74177
## comp 3 25.49091
## comp 4 30.49915
## comp 5 35.24678
## comp 6 39.30810
## comp 7 43.16807
## comp 8 46.70094
## comp 9 50.12997
## comp 10 53.06430
## comp 11 55.76257
## comp 12 58.29264
## comp 13 60.65428
## comp 14 62.82279
## comp 15 64.86832
## comp 16 66.86075
## comp 17 68.77100
## comp 18 70.63561
## comp 19 72.43471
## comp 20 74.20554
## comp 21 75.92309
## comp 22 77.57221
## comp 23 79.10597
## comp 24 80.60078
## comp 25 82.06648
## comp 26 83.42452
## comp 27 84.74007
## comp 28 85.90234
## comp 29 87.03567
## comp 30 88.15155
## comp 31 89.19047
## comp 32 90.20217
## comp 33 91.18208
## comp 34 92.10585
## comp 35 93.01493
## comp 36 93.86473
## comp 37 94.63402
## comp 38 95.39122
## comp 39 96.08908
## comp 40 96.72712
## comp 41 97.33968
## comp 42 97.89177
## comp 43 98.43427
## comp 44 98.83614
## comp 45 99.18534
## comp 46 99.44712
## comp 47 99.70017
## comp 48 99.90691
## comp 49 100.00000
barplot(monACP$eig[,1], main="Eigenvalues", names.arg=1:nrow(monACP$eig))
plot.PCA(monACP)
maDescription = dimdesc(monACP)
head(na.omit(maDescription$Dim.1$quanti), n=10)
## correlation p.value
## et 0.8094060 1.121794e-12
## au 0.7619851 1.298950e-10
## sont 0.7566972 2.061230e-10
## dont 0.7560158 2.185737e-10
## non 0.7325164 1.480523e-09
## moi 0.7301423 1.776392e-09
## droit 0.7223872 3.179344e-09
## uoit 0.7131954 6.183665e-09
## soit 0.7016251 1.378560e-08
## a 0.6906622 2.847610e-08
tail(na.omit(maDescription$Dim.1$quanti), n=10)
## correlation p.value
## co -0.8062675 1.598253e-12
## sun -0.8109536 9.398613e-13
## seit -0.8226655 2.332285e-13
## sunt -0.8349796 4.805490e-14
## e -0.8463918 9.858935e-15
## mei -0.8485535 7.198703e-15
## tut -0.8515295 4.631495e-15
## al -0.8520937 4.255404e-15
## sur -0.8571800 1.951346e-15
## pur -0.8571929 1.947407e-15
head(na.omit(maDescription$Dim.2$quanti), n=10)
## correlation p.value
## mais 0.6675259 1.192037e-07
## car 0.6545848 2.517084e-07
## ains 0.6526491 2.806225e-07
## faire 0.6500554 3.242488e-07
## sans 0.6166221 1.859942e-06
## uenus 0.5942362 5.372931e-06
## sains 0.5907636 6.289294e-06
## tous 0.5895411 6.644928e-06
## dedens 0.5733475 1.348806e-05
## no 0.5678215 1.702859e-05
tail(na.omit(maDescription$Dim.2$quanti), n=10)
## correlation p.value
## es -0.5531734 3.096695e-05
## marchis -0.5573724 2.616361e-05
## seignor -0.5624656 2.126035e-05
## uile -0.5642515 1.975219e-05
## ainz -0.5645785 1.948695e-05
## ge -0.5652377 1.896221e-05
## cite -0.5901032 6.479156e-06
## granz -0.6126725 2.256150e-06
## toz -0.6586748 1.995241e-07
## terre -0.6612321 1.722357e-07
distance = dist(t(monCorpusSelect), method = "manhattan")
2D fit
fit = cmdscale(distance,eig=TRUE, k=2)
x = fit$points[,1]
y = fit$points[,2]
plot(x, y, xlab="Coordinate 1", ylab="Coordinate 2", main="Metric PMD")
text(x, y, labels = row.names(t(monCorpus2)), cex=.7)
isoMDS()
from package MASS
,
library(MASS)
distance = dist(t(monCorpusSelect), method = "manhattan")
fit = isoMDS(distance, k=2)
## initial value 25.144180
## iter 5 value 12.300577
## iter 10 value 10.795732
## iter 15 value 10.583081
## iter 20 value 10.498248
## iter 20 value 10.491333
## iter 20 value 10.486975
## final value 10.486975
## converged
fit
## $points
## [,1] [,2]
## agn_1137pm13_Nord.Ouest_1100caRolS -0.717163253 -0.132077063
## agn_1200pm20_agn_1180pm10_Asprem.P4 -0.889593533 -0.167713855
## agn_1200pm20_nil_nil_Otin_M -0.801173685 -0.155049893
## agn_1213pm13_frc_1125pm25GormB -0.704042608 0.013544142
## agn_1225pm25_agn_1170ca_HornP.C -0.826477551 0.069149707
## agn_1250pm10_agn_1150pm16_ChGuillM -0.613041381 0.016542198
## agn_1250pm10_agn_1170ca_HornP.O -0.952700263 0.002378254
## agn_1250pm16_agn_1180pm10_Asprem.C -2.030447378 0.323354388
## agn_1250pm50_nil_1250pm50_MacaireAl2B -0.428783003 0.905236508
## agn_1275pm25_nil_nil_Otin_B -0.579819885 -0.051937045
## agn_1290pm10_agn_1175pm25_PelCharlB -0.604188822 -0.143024411
## agn_1290pm10_agn_1250pm10_DestrRomeF2 0.004224903 0.614021871
## agn_1335ca_agn_1190pm10_AmAmOctF -0.651959596 0.606195389
## art_1295_nil_1150pm17CharroiSch_C. 0.276792114 -0.050592195
## art_1295_picmérid_1180ca_MonGuill1C2 0.217413661 0.057993529
## bourg_1270ca_champmérid_1210pm10GirVianeE 0.202830787 -0.121547712
## bourg_1270ca_frc_1210pm10GuibAndrM 0.267157782 -0.271364724
## bourg_1270ca_nil_1210pm10_AimeriD 0.207651479 -0.097138660
## bourg_1270ca_nil_1213pm13_MortAymC 0.181084547 -0.135849209
## bourg_1325pm25_Sud.Est_1190pm10_FloovG 0.333955174 -0.189370472
## Est_1300ca_pic_1213pm13FlorenceW 0.154165886 -0.182222316
## frc_1262pm13_nil_1150pm16_CourLouisLe 0.163361490 -0.061951169
## frc_1263pm13_nil_1150pm17CharroiSch_A1. 0.241831664 -0.132303387
## frc_1263pm13_nil_1150pm17CharroiSch_A2. 0.243530244 -0.109813574
## frc_1283pm17_nil_1150pm17CharroiSch_A4. 0.275778571 -0.117480171
## lorr_1290pm10_picmérid_1225caOrsonP 0.450154207 -0.002051281
## lorr_1325pm25_nil_1150pm17CharroiSch_A3. 0.245895134 -0.114880572
## lorr_1387pm13_pic_1365ca_BaudSebC 0.336042051 0.178384976
## lorrmérid_1275pm25_nil_1150pm20CharroiSch_D. 0.414494174 -0.182833306
## lorrsept_1275pm25_nil_1200ca_AmAmD 0.239391515 -0.001920170
## Meuse_1262pm13_lorr_1200caPriseCordD 0.429828289 -0.159017219
## nil_1250pm50_nil_1150pm17CharroiSch_fragm. 0.567486577 -0.448378244
## nil_1250pm50_nil_1230ca_GuiBourgG 0.259275707 -0.018138285
## nil_1300ca_norm_1200caAyeB 0.012295607 0.127302749
## nil_1450pm10_pic_1300ca_EnfGarB. 0.505854066 0.375177155
## Nord_1275pm25_Nord.Est_1190pm10RCambr2M 0.351826350 0.037832872
## Nord.Est_1262pm13_Nord.Est_1190pm10_PriseOrabR1 0.169026131 -0.085947909
## Paris_1290pm10_flandr_1275_AdenBuevH 0.267719062 0.141956131
## Paris_1335ca_nil_1150pm17CharroiSch_B1. 0.290015337 -0.029222630
## Paris_1335ca_nil_1150pm17CharroiSch_B2. 0.277154316 0.005534127
## pic_1213pm13_pic_1190pm10_AliscW 0.158642898 0.001597015
## pic_1213pm13_pic_1190pm10_MonRaincB 0.196106268 0.049105058
## pic_1213pm13_picmérid_1150pm16_MonGuill1C1 0.223333378 0.085734106
## pic_1225pm25_Nord.Est_1190pm10RCambr1M 0.343512131 -0.022699646
## pic_1275pm25_pic_1160caAiol1N 0.186744109 0.032048993
## pic_1275pm25_pic_1190pm10ElieB. 0.219088689 0.032931241
## pic_1275pm25_pic_1210pm10_Aiol2N 0.189949632 0.021175611
## pic_1290pm10_pic_1275pm25FlorOctOctV 0.388957773 0.139304754
## StBrieuc_1317_nil_1190ca_Fier.V 0.169470221 -0.595434176
## StBrieuc_1317_nil_nil_Otin_A 0.137349034 -0.056541479
##
## $stress
## [1] 10.48698
x = fit$points[,1]
y = fit$points[,2]
plot(x, y, xlab="Coordonnée 1", ylab="Coordonnée 1",
main="Non metric PMD", type="n")
text(x, y, labels = row.names(t(monCorpus2)), cex=.7)
monCorpusSelect = monCorpus2[1:1000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
monCorpusSelect = monCorpus2[1:1000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
monCorpusSelect = monCorpus2[1:1200,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
monCorpusSelect = monCorpus2[1:2000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
monCorpusSelect = monCorpus2[1:3000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
monCorpusSelect = monCorpus2[1:4000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
Based on the deformation they seem to cause to previous CAH, let’s remove texts with less than 2000 words:
[49] “StBrieuc_1317_nil_1190ca_Fier.V” , 1682 [9] “agn_1250pm50_nil_1250pm50_MacaireAl2B”, 1179 [32] “nil_1250pm50_nil_1150pm17CharroiSch_fragm.” , 879 [8] “agn_1250pm16_agn_1180pm10_Asprem.C”, 387
monCorpus3 = monCorpus2[,-c(49,9,32,8)]
boxplot(colSums(monCorpus3), main = "Number of words per text", ylab="Number of words", sub=nomCorpus)
#Re-sorting words
monCorpus3 = monCorpus3[order(rowSums(monCorpus3), decreasing = TRUE),]
monCorpusSelect = monCorpus3[1:1000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
It is then possible to separate witnesses in different classes, and compute their specificities. Nb classes can be chosen with the help of a height plot.
CAH2 = as.hclust(CAH)
plot(CAH2$height, type="h", ylab="hauteurs")
We can then describe the classes
classes = cutree(CAH, k = "3")
#Adding classes to the table
monCorpusAvecClasses = t(monCorpusSelect)
monCorpusAvecClasses = cbind(as.data.frame(monCorpusAvecClasses), as.factor(classes))
colnames(monCorpusAvecClasses[ncol(monCorpusAvecClasses)]) = "Classes"
#And describing
library(FactoMineR)
mesClasses = catdes(monCorpusAvecClasses, num.var = ncol(monCorpusAvecClasses))
#Looking in more detail at some descriptions
head(mesClasses$quanti$`1`, n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## pur 6.150242 0.007304998 0.0017474436 0.002774797 0.003398287
## sunt 6.030072 0.006376563 0.0016132262 0.002586779 0.002970690
## ad 5.911688 0.013054791 0.0031299427 0.006101124 0.006313656
## mei 5.829337 0.002082077 0.0004978879 0.001034190 0.001022013
## sur 5.794786 0.004868630 0.0012296727 0.002363936 0.002361610
## tut 5.755748 0.004923835 0.0012118677 0.002532150 0.002425330
## lur 5.746153 0.004366352 0.0010441277 0.002294333 0.002174305
## al 5.622060 0.007859435 0.0022036718 0.003694655 0.003783244
## e 5.602756 0.039127887 0.0108415916 0.014023607 0.018986414
## sun 5.560295 0.007618652 0.0018262512 0.004481422 0.003917683
## p.value
## pur 7.736496e-10
## sunt 1.638864e-09
## ad 3.386185e-09
## mei 5.564801e-09
## sur 6.840829e-09
## tut 8.625893e-09
## lur 9.129686e-09
## al 1.886935e-08
## e 2.109702e-08
## sun 2.693184e-08
tail(mesClasses$quanti$`1`, n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## droit -4.873678 6.648641e-05 0.0009769109 0.0002102485 0.0007025151
## uoit -4.896590 3.324321e-05 0.0012831772 0.0001051243 0.0009599793
## dont -4.969360 1.181674e-04 0.0020457297 0.0001951748 0.0014587356
## au -4.973441 8.223470e-04 0.0045650893 0.0021521208 0.0028300984
## non -4.998135 5.452789e-05 0.0009564187 0.0001406521 0.0006786010
## a -5.011458 2.695149e-02 0.0361933395 0.0058019338 0.0069352730
## moi -5.061471 2.776890e-04 0.0019990380 0.0004756086 0.0012789710
## mon -5.147234 3.433947e-04 0.0025590944 0.0007120460 0.0016188461
## et -5.161155 1.012365e-02 0.0391548996 0.0212841472 0.0211537543
## sont -5.220152 3.213510e-04 0.0031190234 0.0010162011 0.0020154977
## p.value
## droit 1.095392e-06
## uoit 9.751378e-07
## dont 6.717436e-07
## au 6.577480e-07
## non 5.788735e-07
## a 5.401913e-07
## moi 4.160331e-07
## mon 2.643553e-07
## et 2.454308e-07
## sont 1.787765e-07
head(mesClasses$quanti$`2`, n=10)
## v.test Mean in category Overall mean sd in category
## toz 4.474021 0.0017023915 0.0008434564 0.0012403362
## tot 4.382810 0.0033660192 0.0019674502 0.0017741771
## rois 4.310599 0.0044414237 0.0028822382 0.0017626344
## cheualier 4.261726 0.0024184910 0.0015919849 0.0007987976
## roi 4.255214 0.0028503275 0.0018160613 0.0012184779
## uos 4.178890 0.0122224852 0.0079625916 0.0053207325
## desus 4.164201 0.0005132772 0.0002884785 0.0003352833
## moi 4.144333 0.0028611614 0.0019990380 0.0008756075
## auoir 3.993397 0.0014021988 0.0009233315 0.0006385095
## auferrant 3.969725 0.0002854385 0.0001533879 0.0002217018
## Overall sd p.value
## toz 0.0011803430 7.676223e-06
## tot 0.0019619011 1.171581e-05
## rois 0.0022238529 1.628129e-05
## cheualier 0.0011923573 2.028539e-05
## roi 0.0014943654 2.088489e-05
## uos 0.0062673464 2.929352e-05
## desus 0.0003319006 3.124449e-05
## moi 0.0012789710 3.408040e-05
## auoir 0.0007372564 6.513332e-05
## auferrant 0.0002045153 7.195560e-05
tail(mesClasses$quanti$`2`, n=10)
## v.test Mean in category Overall mean sd in category
## sur -3.053893 5.662408e-05 0.0012296727 2.099547e-04
## nel -3.059150 3.204821e-04 0.0008494206 5.990498e-04
## del -3.077019 1.659918e-03 0.0027140111 1.587501e-03
## u -3.091304 2.637333e-04 0.0012274229 4.139493e-04
## pur -3.161469 0.000000e+00 0.0017474436 0.000000e+00
## ensamble -3.248262 2.075725e-05 0.0001591343 5.277235e-05
## mais -3.477778 1.240853e-03 0.0026230608 1.948339e-03
## al -3.495605 5.266973e-05 0.0022036718 1.028980e-04
## faire -3.581903 3.022582e-04 0.0005825763 3.780505e-04
## fait -3.880163 1.325204e-03 0.0027574043 1.621045e-03
## Overall sd p.value
## sur 0.0023616097 0.0022589279
## nel 0.0010630417 0.0022196617
## del 0.0021061754 0.0020908180
## u 0.0019166433 0.0019927940
## pur 0.0033982874 0.0015697542
## ensamble 0.0002619142 0.0011611237
## mais 0.0024435279 0.0005055891
## al 0.0037832437 0.0004729875
## faire 0.0004811530 0.0003411006
## fait 0.0022693406 0.0001043866
head(mesClasses$quanti$`3`, n=10)
## v.test Mean in category Overall mean sd in category
## ains 5.408400 0.0015253768 5.612932e-04 0.0006529045
## sains 5.082391 0.0004500108 1.809083e-04 0.0002331815
## tous 4.955749 0.0018226451 6.847528e-04 0.0009751956
## dedens 4.906695 0.0009463731 3.917381e-04 0.0003556854
## sans 4.779205 0.0015055952 6.757197e-04 0.0007416895
## trestous 4.653482 0.0002954394 1.152374e-04 0.0001866660
## tout 4.650212 0.0038803997 1.672939e-03 0.0016585782
## toutes 4.603120 0.0003199332 1.311994e-04 0.0001806836
## dolans 4.536534 0.0002494829 8.494561e-05 0.0002004730
## maistre 4.533244 0.0004088393 1.725775e-04 0.0001854301
## Overall sd p.value
## ains 0.0007909356 6.359024e-08
## sains 0.0002349336 3.727140e-07
## tous 0.0010187958 7.205207e-07
## dedens 0.0005015491 9.262377e-07
## sans 0.0007704645 1.759900e-06
## trestous 0.0001718212 3.263754e-06
## tout 0.0021062773 3.315937e-06
## toutes 0.0001819252 4.162078e-06
## dolans 0.0001609294 5.718618e-06
## maistre 0.0002312490 5.808472e-06
tail(mesClasses$quanti$`3`, n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## escuz -3.146018 1.119065e-05 0.0002518119 3.211569e-05 0.0003393658
## fet -3.159369 5.709486e-05 0.0014979408 1.398930e-04 0.0020235434
## tens -3.183511 9.193638e-06 0.0002409550 2.556520e-05 0.0003230205
## anz -3.194625 1.059614e-05 0.0002285938 3.820492e-05 0.0003027801
## auez -3.230122 4.120227e-04 0.0011750134 6.159440e-04 0.0010480829
## piez -3.305240 1.010120e-05 0.0003784395 2.108560e-05 0.0004944692
## sanz -3.354954 1.376751e-06 0.0005358283 4.963945e-06 0.0007068335
## bone -3.688227 1.798580e-04 0.0005903337 2.481718e-04 0.0004938160
## mes -3.740337 2.230568e-03 0.0043820954 7.226719e-04 0.0025522979
## granz -3.747034 6.883753e-06 0.0007540766 2.481972e-05 0.0008847902
## p.value
## escuz 0.0016550987
## fet 0.0015811133
## tens 0.0014550083
## anz 0.0014001265
## auez 0.0012373760
## piez 0.0009489526
## sanz 0.0007937827
## bone 0.0002258223
## mes 0.0001837736
## granz 0.0001789378
plot(mesClasses)
monCorpusSelect = monCorpus3[1:2000,]
for(i in 1:ncol(monCorpusSelect)){
monCorpusSelect[,i] = monCorpusSelect[,i]/sum(monCorpusSelect[,i])
}
library(cluster)
CAH = agnes(t(monCorpusSelect), metric="manhattan", method = "ward")
plot(CAH, which.plots = 2, main = "CAH", xlab=paste(nrow(monCorpusSelect), " MFW -- Manhattan dist."))
It is then possible to separate witnesses in different classes, and compute their specificities. Nb classes can be chosen with the help of a height plot.
CAH2 = as.hclust(CAH)
plot(CAH2$height, type="h", ylab="hauteurs")
We can then describe the classes
classes = cutree(CAH, k = "3")
#Adding classes to the table
monCorpusAvecClasses = t(monCorpusSelect)
monCorpusAvecClasses = cbind(as.data.frame(monCorpusAvecClasses), as.factor(classes))
colnames(monCorpusAvecClasses[ncol(monCorpusAvecClasses)]) = "Classes"
#And describing
library(FactoMineR)
mesClasses = catdes(monCorpusAvecClasses, num.var = ncol(monCorpusAvecClasses))
#Looking in more detail at some descriptions
head(round(mesClasses$quanti$`1`, digits=4), n=25)
## v.test Mean in category Overall mean sd in category Overall sd
## pur 6.1376 0.0067 0.0016 0.0026 0.0031
## sunt 6.0154 0.0058 0.0015 0.0024 0.0027
## ad 5.9070 0.0120 0.0029 0.0056 0.0058
## mei 5.8200 0.0019 0.0005 0.0010 0.0009
## sur 5.7982 0.0044 0.0011 0.0021 0.0022
## lur 5.7501 0.0040 0.0010 0.0021 0.0020
## tut 5.7371 0.0045 0.0011 0.0023 0.0022
## al 5.6221 0.0072 0.0020 0.0034 0.0035
## e 5.6007 0.0357 0.0099 0.0127 0.0173
## sun 5.5464 0.0070 0.0017 0.0041 0.0036
## seit 5.4748 0.0020 0.0005 0.0012 0.0010
## dunt 5.4731 0.0018 0.0004 0.0011 0.0010
## od 5.4573 0.0033 0.0008 0.0019 0.0017
## mun 5.3219 0.0018 0.0004 0.0012 0.0010
## si 5.3008 0.0186 0.0135 0.0030 0.0037
## funt 5.2749 0.0008 0.0002 0.0006 0.0004
## reis 5.1917 0.0046 0.0011 0.0033 0.0025
## seignurs 5.1741 0.0009 0.0002 0.0006 0.0005
## rei 5.1565 0.0038 0.0009 0.0027 0.0021
## uus 5.0396 0.0066 0.0016 0.0050 0.0037
## unt 4.9906 0.0023 0.0006 0.0018 0.0013
## co 4.9686 0.0036 0.0009 0.0029 0.0021
## nun 4.9614 0.0008 0.0002 0.0006 0.0005
## u 4.9366 0.0034 0.0011 0.0020 0.0018
## io 4.9246 0.0038 0.0009 0.0030 0.0022
## p.value
## pur 0
## sunt 0
## ad 0
## mei 0
## sur 0
## lur 0
## tut 0
## al 0
## e 0
## sun 0
## seit 0
## dunt 0
## od 0
## mun 0
## si 0
## funt 0
## reis 0
## seignurs 0
## rei 0
## uus 0
## unt 0
## co 0
## nun 0
## u 0
## io 0
tail(round(mesClasses$quanti$`1`, digits=4), n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## droit -4.8741 0.0001 0.0009 0.0002 0.0006
## uoit -4.8998 0.0000 0.0012 0.0001 0.0009
## dont -4.9669 0.0001 0.0019 0.0002 0.0013
## au -4.9835 0.0008 0.0042 0.0020 0.0026
## non -5.0134 0.0000 0.0009 0.0001 0.0006
## moi -5.0723 0.0003 0.0018 0.0004 0.0012
## a -5.0781 0.0246 0.0332 0.0050 0.0064
## mon -5.1568 0.0003 0.0023 0.0006 0.0015
## et -5.1677 0.0093 0.0359 0.0195 0.0194
## sont -5.2137 0.0003 0.0029 0.0009 0.0019
## p.value
## droit 0
## uoit 0
## dont 0
## au 0
## non 0
## moi 0
## a 0
## mon 0
## et 0
## sont 0
head(mesClasses$quanti$`2`, n=10)
## v.test Mean in category Overall mean sd in category
## toz 4.478658 0.0015561647 0.0007710472 0.0011309974
## tot 4.375692 0.0030835690 0.0018029717 0.0016315537
## cuit 4.372575 0.0002490371 0.0001278660 0.0001834762
## rois 4.291429 0.0040649230 0.0026419057 0.0016192577
## cheualier 4.238178 0.0022120757 0.0014589207 0.0007325382
## roi 4.233783 0.0026088945 0.0016647435 0.0011215995
## uos 4.169151 0.0111887216 0.0072947110 0.0048826947
## desus 4.144129 0.0004691662 0.0002643345 0.0003066657
## moi 4.122455 0.0026166624 0.0018318164 0.0007996889
## auoir 3.965095 0.0012840977 0.0008470702 0.0005895703
## Overall sd p.value
## toz 0.0010777864 7.511382e-06
## tot 0.0017993337 1.210478e-05
## cuit 0.0001703758 1.227895e-05
## rois 0.0020387040 1.775273e-05
## cheualier 0.0010925746 2.253412e-05
## roi 0.0013710670 2.297922e-05
## uos 0.0057424260 3.057365e-05
## desus 0.0003038854 3.411071e-05
## moi 0.0011705082 3.748557e-05
## auoir 0.0006776432 7.336679e-05
tail(mesClasses$quanti$`2`, n=10)
## v.test Mean in category Overall mean sd in category
## nel -3.071761 2.921670e-04 7.762399e-04 5.459934e-04
## u -3.086083 2.408151e-04 1.125502e-03 3.778036e-04
## del -3.094420 1.515588e-03 2.484454e-03 1.446934e-03
## pur -3.154992 0.000000e+00 1.599425e-03 0.000000e+00
## ensamble -3.249854 1.912161e-05 1.463373e-04 4.869693e-05
## mais -3.481179 1.136516e-03 2.409430e-03 1.780200e-03
## al -3.499477 4.820102e-05 2.016114e-03 9.416327e-05
## faire -3.585149 2.769231e-04 5.351283e-04 3.459125e-04
## fuissent -3.608004 2.100995e-06 5.431677e-05 9.395934e-06
## fait -3.890525 1.216125e-03 2.525689e-03 1.495133e-03
## Overall sd p.value
## nel 9.688784e-04 0.0021280009
## u 1.762494e-03 0.0020281193
## del 1.924999e-03 0.0019719842
## pur 3.116819e-03 0.0016050234
## ensamble 2.406704e-04 0.0011546411
## mais 2.248115e-03 0.0004992124
## al 3.457391e-03 0.0004661715
## faire 4.427961e-04 0.0003368857
## fuissent 8.897759e-05 0.0003085616
## fait 2.069495e-03 0.0001000275
head(mesClasses$quanti$`3`, n=40)
## v.test Mean in category Overall mean sd in category
## ains 5.409832 1.406578e-03 5.172708e-04 6.031454e-04
## sains 5.089362 4.139562e-04 1.663530e-04 2.136303e-04
## passes 5.028743 1.565728e-04 5.026232e-05 1.080122e-04
## tous 4.968268 1.680063e-03 6.301646e-04 8.981395e-04
## dedens 4.919486 8.718890e-04 3.603755e-04 3.276282e-04
## laissa 4.858387 1.100498e-04 3.626568e-05 8.078015e-05
## commanda 4.784853 1.045703e-04 3.182573e-05 8.570037e-05
## sans 4.779803 1.387637e-03 6.215808e-04 6.861405e-04
## tout 4.670654 3.574118e-03 1.536767e-03 1.528719e-03
## trestous 4.661928 2.724132e-04 1.060404e-04 1.723743e-04
## toutes 4.613347 2.945913e-04 1.206043e-04 1.664115e-04
## maistre 4.550576 3.766488e-04 1.586368e-04 1.709456e-04
## mais 4.538221 4.708800e-03 2.409430e-03 1.496869e-03
## dolans 4.534993 2.297414e-04 7.820176e-05 1.848393e-04
## toute 4.513328 8.154328e-04 3.837880e-04 3.660893e-04
## sarrasins 4.482699 3.546568e-04 1.170169e-04 2.975713e-04
## ensamble 4.468342 3.887045e-04 1.463373e-04 2.216853e-04
## soies 4.404860 3.501856e-04 1.394567e-04 2.109537e-04
## chou 4.370183 6.520479e-04 2.009220e-04 6.291789e-04
## maris 4.317473 1.186399e-04 4.012836e-05 1.008126e-04
## mieus 4.294786 2.845108e-04 8.966726e-05 2.778141e-04
## doiuent 4.278769 8.463444e-05 3.253332e-05 5.642585e-05
## no 4.266644 4.270914e-04 1.575664e-04 3.657351e-04
## solaus 4.247424 9.440792e-05 3.040047e-05 9.161727e-05
## lieu 4.233871 1.673524e-04 5.344567e-05 1.650636e-04
## lies 4.222888 2.678242e-04 1.043338e-04 1.740069e-04
## haus 4.222000 8.447641e-05 2.737784e-05 8.208264e-05
## entres 4.220293 2.098408e-04 7.636474e-05 1.549809e-04
## cousin 4.200350 1.110112e-04 3.847578e-05 9.720733e-05
## menes 4.192784 1.139253e-04 4.203144e-05 8.525681e-05
## dieu 4.124579 2.674281e-03 1.345647e-03 1.223182e-03
## ainc 4.121052 6.299958e-04 2.343479e-04 5.384821e-04
## faire 4.112571 9.455421e-04 5.351283e-04 2.863789e-04
## desous 4.110031 2.646247e-04 9.598003e-05 2.469042e-04
## cha 4.103084 4.126646e-04 1.280662e-04 4.403300e-04
## biaus 4.096747 8.714449e-04 4.087061e-04 4.963774e-04
## espiel 4.096126 2.880943e-04 8.996775e-05 3.068360e-04
## mors 4.093211 7.086827e-04 3.406439e-04 2.702706e-04
## montes 4.081574 2.929712e-04 1.109115e-04 2.301039e-04
## tenrement 4.076297 6.651975e-05 2.024514e-05 7.251305e-05
## Overall sd p.value
## ains 7.293959e-04 6.308394e-08
## sains 2.158680e-04 3.592696e-07
## passes 9.380193e-05 4.937058e-07
## tous 9.376430e-04 6.755358e-07
## dedens 4.613522e-04 8.677165e-07
## laissa 6.738547e-05 1.183459e-06
## commanda 6.745700e-05 1.711125e-06
## sans 7.111248e-04 1.754672e-06
## tout 1.935456e-03 3.002426e-06
## trestous 1.583478e-04 3.132606e-06
## toutes 1.673385e-04 3.962365e-06
## maistre 2.125736e-04 5.349926e-06
## mais 2.248115e-03 5.673074e-06
## dolans 1.482672e-04 5.760531e-06
## toute 4.243507e-04 6.381831e-06
## sarrasins 2.352205e-04 7.370488e-06
## ensamble 2.406704e-04 7.882819e-06
## soies 2.122693e-04 1.058522e-05
## chou 4.580295e-04 1.241425e-05
## maris 8.068621e-05 1.578258e-05
## mieus 2.012982e-04 1.748622e-05
## doiuent 5.402857e-05 1.879298e-05
## no 2.802902e-04 1.984355e-05
## solaus 6.686521e-05 2.162424e-05
## lieu 1.193733e-04 2.297027e-05
## lies 1.717821e-04 2.411913e-05
## haus 6.000706e-05 2.421439e-05
## entres 1.403318e-04 2.439849e-05
## cousin 7.662311e-05 2.665022e-05
## menes 7.608247e-05 2.755520e-05
## dieu 1.429292e-03 3.714136e-05
## ainc 4.259867e-04 3.771465e-05
## faire 4.427961e-04 3.912775e-05
## desous 1.820634e-04 3.956053e-05
## cha 3.077637e-04 4.076786e-05
## biaus 5.011781e-04 4.189956e-05
## espiel 2.146173e-04 4.201219e-05
## mors 3.989557e-04 4.254394e-05
## montes 1.979163e-04 4.473177e-05
## tenrement 5.037003e-05 4.575857e-05
tail(mesClasses$quanti$`3`, n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## escuz -3.152697 1.032201e-05 0.0002292556 2.958106e-05 0.0003081239
## fet -3.154568 5.279559e-05 0.0013693657 1.294578e-04 0.0018518226
## tens -3.181822 8.496953e-06 0.0002204283 2.365666e-05 0.0002955389
## anz -3.190269 9.782726e-06 0.0002090418 3.527212e-05 0.0002771317
## auez -3.217091 3.788943e-04 0.0010758081 5.647197e-04 0.0009611938
## piez -3.317069 9.323632e-06 0.0003448047 1.946336e-05 0.0004487545
## sanz -3.354431 1.274714e-06 0.0004892993 4.596045e-06 0.0006455326
## bone -3.687228 1.660404e-04 0.0005397281 2.291591e-04 0.0004496805
## mes -3.726697 2.054225e-03 0.0040114798 6.648747e-04 0.0023303363
## granz -3.742516 6.373568e-06 0.0006888167 2.298022e-05 0.0008090922
## p.value
## escuz 0.0016176971
## fet 0.0016073588
## tens 0.0014635196
## anz 0.0014214030
## auez 0.0012949740
## piez 0.0009096720
## sanz 0.0007952850
## bone 0.0002267101
## mes 0.0001940058
## granz 0.0001821867
plot(mesClasses)
To describe the Picard group without the artifact of the Northern Lotharingian subgroup, that did not appear in the orginal results (see Camps, 2016):
Group 4, without [28] “lorrsept_1275pm25_nil_1200ca_AmAmD” [30] “nil_1250pm50_nil_1230ca_GuiBourgG” [33] “Nord_1275pm25_Nord.Est_1190pm10RCambr2M” [41] “pic_1225pm25_Nord.Est_1190pm10RCambr1M”
monCorpusAvecClasses2 = monCorpusAvecClasses[-c(28,30,33,41),]
mesClasses = catdes(monCorpusAvecClasses2, num.var = ncol(monCorpusAvecClasses))
#Looking in more detail at some descriptions
head(round(mesClasses$quanti$`3`, digits=4), n=25)
## v.test Mean in category Overall mean sd in category Overall sd
## ains 5.6322 0.0016 0.0005 0.0003 0.0007
## tous 5.4891 0.0021 0.0006 0.0006 0.0010
## passes 5.2743 0.0002 0.0000 0.0001 0.0001
## chou 5.2216 0.0009 0.0002 0.0006 0.0005
## trestous 5.0875 0.0003 0.0001 0.0001 0.0002
## tout 5.0120 0.0043 0.0015 0.0010 0.0020
## sarrasins 4.9654 0.0004 0.0001 0.0003 0.0002
## sains 4.9536 0.0004 0.0001 0.0002 0.0002
## toutes 4.9496 0.0004 0.0001 0.0001 0.0002
## commanda 4.9074 0.0001 0.0000 0.0001 0.0001
## cha 4.9023 0.0006 0.0001 0.0004 0.0003
## mieus 4.8405 0.0004 0.0001 0.0003 0.0002
## ochis 4.7118 0.0002 0.0000 0.0002 0.0001
## no 4.6579 0.0005 0.0002 0.0004 0.0003
## lieu 4.6264 0.0002 0.0001 0.0002 0.0001
## uausist 4.6239 0.0002 0.0000 0.0001 0.0001
## espiel 4.6180 0.0004 0.0001 0.0003 0.0002
## laissa 4.6063 0.0001 0.0000 0.0001 0.0001
## dolans 4.5675 0.0003 0.0001 0.0002 0.0002
## chi 4.5667 0.0009 0.0003 0.0006 0.0005
## toute 4.5588 0.0009 0.0004 0.0002 0.0004
## cief 4.4868 0.0007 0.0002 0.0005 0.0004
## ainc 4.4662 0.0008 0.0002 0.0005 0.0004
## mais 4.4656 0.0052 0.0023 0.0014 0.0023
## ceual 4.4543 0.0006 0.0002 0.0005 0.0003
## p.value
## ains 0
## tous 0
## passes 0
## chou 0
## trestous 0
## tout 0
## sarrasins 0
## sains 0
## toutes 0
## commanda 0
## cha 0
## mieus 0
## ochis 0
## no 0
## lieu 0
## uausist 0
## espiel 0
## laissa 0
## dolans 0
## chi 0
## toute 0
## cief 0
## ainc 0
## mais 0
## ceual 0
tail(round(mesClasses$quanti$`3`, digits=4), n=10)
## v.test Mean in category Overall mean sd in category Overall sd
## donez -2.9366 0e+00 0.0002 0e+00 0.0002
## ainz -2.9825 0e+00 0.0009 1e-04 0.0010
## ci -3.0037 4e-04 0.0011 4e-04 0.0009
## tuit -3.0366 1e-04 0.0006 2e-04 0.0006
## deu -3.0514 1e-04 0.0016 2e-04 0.0017
## armez -3.0935 0e+00 0.0002 0e+00 0.0002
## auez -3.2057 2e-04 0.0011 4e-04 0.0010
## granz -3.2662 0e+00 0.0008 0e+00 0.0008
## mes -3.2776 2e-03 0.0042 8e-04 0.0024
## bone -3.3023 2e-04 0.0006 2e-04 0.0005
## p.value
## donez 0.0033
## ainz 0.0029
## ci 0.0027
## tuit 0.0024
## deu 0.0023
## armez 0.0020
## auez 0.0013
## granz 0.0011
## mes 0.0010
## bone 0.0010