CODEWITHSUNDEEP
r
drmariod
drmariod

Reputation: 11772

using t.test with apply on some columns of a data frame

I am trying to calculate a p-value for some specific columns in a data frame.

df <- data.frame(
  expa_1=sample(c(rnorm(40,5,1) + 30)),
  expa_2=sample(c(rnorm(40,5,1) + 30)),
  expa_3=sample(c(rnorm(40,5,1) + 30)),
  expb_1=sample(c(rnorm(40,15,1) + 30)),
  expb_2=sample(c(rnorm(40,15,1) + 30)),
  expb_3=sample(c(rnorm(40,15,1) + 30)))
apply(df,1,function(x) t.test(x[1:3],x[4:6])$p.value)

So this is my little example which seems to work fine... If I have more than just the columns I want to calculate the p-value from, then I get an error.

df <- data.frame(names=c(letters,LETTERS[1:14]),
                 expa_1=sample(c(rnorm(40,5,1) + 30)),
                 expa_2=sample(c(rnorm(40,5,1) + 30)),
                 expa_3=sample(c(rnorm(40,5,1) + 30)),
                 expb_1=sample(c(rnorm(40,15,1) + 30)),
                 expb_2=sample(c(rnorm(40,15,1) + 30)),
                 expb_3=sample(c(rnorm(40,15,1) + 30)))
apply(df,1,function(x) t.test(x[2:4],x[5:7])$p.value)

Can somebody explain this to me and maybe give me a better solution for this?

Thanks in advance

Upvotes: 3

Views: 659

Answers (1)

akrun
akrun

Reputation: 887831

You could try

 apply(df[-1], 1, function(x) t.test(x[1:3], x[4:6])$p.value)

The first column is character/factor. When you do apply, it converts to matrix and the values will be converted to 'character' as 'matrix' can hold only one class.

For example

 apply(df,1,function(x) x[2:4])[1:3]
#[1] "34.33018" "34.93256" "34.74045"

Upvotes: 4

Related Questions