CODEWITHSUNDEEP
rp-value
albert
albert

Reputation: 325

Manipulating the value of p-value of output in R

Suppose

data11 <- c(0.5388417, 0.7263466, 0.3612457, 0.2495263, 0.1780413)
data22 <- c(0.674262245, 0.659560230, 0.001548212, 0.501228052, 0.802885484)
d=as.data.frame(cbind(dat=c(data11,data22),level=c(rep(1,length(data11)),rep(2,length(data22)))))

then, I use the function oneway_test to get tje p-value

library(coin)
oneway_test(d$dat~as.factor(d$level))

I get:

    Asymptotic Two-Sample Fisher-Pitman Permutation Test

data:  d$dat by as.factor(d$level) (1, 2)
Z = -0.70257, p-value = 0.4823
alternative hypothesis: true mu is not equal to 0

But, how can I save the value of P-value? . Try to do this: d$p.value. I do not get it

Thanks

Upvotes: 3

Views: 304

Answers (2)

thc
thc

Reputation: 9705

Check the documentation of the coin package. You can use the pvalue function to return the pvalue:

results <- oneway_test(d$dat~as.factor(d$level))
pval <- pvalue(results)

Upvotes: 2

agenis
agenis

Reputation: 8377

I guess this is not straighforward because the printed result (p-value) is not contained in the returned object of the function oneway_test but is printed as a side effect. You might use some regex as a workaround:

output <- oneway_test(d$dat~as.factor(d$level))

FindPValue = function(output){
  temp <- capture.output(output)
  return(as.numeric(gsub(".*p-value = ", "", temp[5])))
}

FindPValue(output)
#### [1] 0.4823

Thanks to Adam Quek for the advice with capture.output

Upvotes: 1

Related Questions