I ran into two issues and would appreciate your feedback.
Issue 1
In my testing script below, rearranging the dataframe changes the output (Case 1 vs Case 2 and Case 3 vs Case 4). Why is this?
Issue 2
While the estimated p-values between the Bioconductor (Cases 1 and 2) and developer versions (Cases 3 and 4) are similar, I obtain vastly different Benjamini-Hochberg corrected p-values. Why is this? Which should I trust?
Hi there,
I ran into two issues and would appreciate your feedback.
Issue 1
In my testing script below, rearranging the dataframe changes the output (Case 1 vs Case 2 and Case 3 vs Case 4). Why is this?
Issue 2
While the estimated p-values between the Bioconductor (Cases 1 and 2) and developer versions (Cases 3 and 4) are similar, I obtain vastly different Benjamini-Hochberg corrected p-values. Why is this? Which should I trust?
Versions
Testing script
Output
Cases 1-4