#read file
matechoice=read.csv("matechoice_CCL.csv")

#checking whether the empirical correlation is different from random.
matechoice_allsample=matechoice_sample
vals2=NULL
for (i in 1:1000) {
  set.seed(i)
  matechoice_sample=data.frame(Female=matechoice$Female, Male=sample(matechoice$Male))
  #matechoice_allsample=rbind(matechoice_allsample, matechoice_sample)
  vals2=c(vals2, cor(matechoice_sample$Male,matechoice_sample$Female))
}
hist(vals2, breaks=50)
abline(v=cor(matechoice$Male, matechoice$Female), col="red")
X=cor(matechoice$Male, matechoice$Female)
sum(as.numeric(vals2<=X))/length(vals2)