Problem in calculating p values

Question

[anasjamshed]

## pvalues are given to retrieve the genes which pass the p values
for (i in c(0.01, 0.05, 0.001, 1e-04, 1e-05, 1e-06, 1e-07)) print(paste("genes with p-values smaller than", 
    i, length(which(pvals < i))))

when I try to run this script the following error comes :

Error in which(pvals < i) : object 'pvals' not found

help to overcome this problem

Answer 1

[Kevin Blighe]

You have no variable called pvals in your workspace. Go back to the beginning of your code and determine why.

Thank you!

Kevin

[anasjamshed]

brother pvals is a function to calculate p value its not variable

[Kevin Blighe]

You sure dude? In your code, you are passing it as the first argument to the which() function

[swbarnes2]

The way you have written your code, pvals is a variable.

Functions need inputs. They don't just sit there as barewords.

[anasjamshed]

sorry I already declared the variable after making t-test function

ttest = 
    function(x) {
        tt = t.test(x[logX.cl == 0], x[logX.cl == 1])
        return(c(tt$statistic, tt$p.value))
    }

ans = apply(logX, 1, ttest)
ts <- ans[1, ]
pvals <- ans[2, ]                                                                                                                                                                                       
for (i in c(0.01, 0.05, 0.001, 1e-04, 1e-05, 1e-06, 1e-07)) print(paste("genes with p-values smaller than", 
    i, length(which(pvals < i))))

[Kevin Blighe]

What you're trying to do I would have done a different way, i.e., using Limma (as I know that you are working on microarray data). After that, simple subset() commands on the Limma output could get you want.

If you want to continue with your current code, then:

• check the output of every command and become familiar with all variables in your workspace
• try not to blindly run for loops without first testing a few passes (by setting pre-specified values of i, in this case)
• add message() commands to your loop, too, so that you can verify and QC check each pass.

[zx8754]

Functions do not always need inputs: foo <- function(){ 1+1 } or Sys.Date(), they could also sit as "barewords": do.call(rbind, mydata).

But, yes, agree, in this case pvals is a variable.