CODEWITHSUNDEEP
r
user1971455
user1971455

Reputation: 433

How to apply wilcox.test to compare every column of table

I have a data table in r called tbl. I want to perform a wilcox test to compare every column with every other column in the table and store the p-values in a new table or matrix.

To compare a single pair of columns (1 and 2 in this case) I would normally just do this:

wilcox.test(tbl[,1],tbl[,2],alternative="t",paired=FALSE)

Any help is greatly appreciated.

Upvotes: 4

Views: 1955

Answers (1)

LyzandeR
LyzandeR

Reputation: 37889

One way is to store the 2-way combinations of your tbl and then use lapply to calculate the p.values. Something like this:

#example data.frame
df <- data.frame(a = runif(10), b= runif(10), c=runif(10))

#calculate combinations of names
combinations <- as.data.frame(combn(names(df), 2), stringsAsFactors=FALSE)

#use the above combinations and calculate the wilcox.test to get the p.values
lapply(combinations, function(x) {
  wilcox.test(df[,x[1]] , df[, x[2]], alternative="t", paired=FALSE)$p.value
})

Output:

$V1
[1] 0.8534283

$V2
[1] 0.2175626

$V3
[1] 0.3526814

Upvotes: 8

Related Questions