0
Vote

FastLmm continues after filtering out all SNPs as constants and fails later in MKL with a confusing message about bad parameters.

description

When there is a bug in my data files and 0 individuals remain after intersection, I don’t get an error about that, but instead end up getting a downstream error, a few minutes later, after being told that each of my 500K SNPs has no variation. 
It would be nice if it failed up front with “0 individuals remain after intersection”.

comments

wrote Apr 18, 2012 at 12:01 AM

JennL email:
I noticed that when there is a bug in my data files and 0 individuals remain after intersection, I don’t get an error about that, but instead end up getting a downstream error, a few minutes later, after being told that each of my 500K SNPs has no variation.  Would be nice if it just failed up front with “0 individuals remain after intersection”.
 

wrote Apr 18, 2012 at 12:03 AM

This appears to be related to binary input files only.  Other filetypes do fail with a message about no SNPs passing the filter

wrote Feb 13, 2013 at 11:40 PM