check_sumstat_format failed (likely due to leading space in the first column)

[ofrei]

I'm getting the following error

My input has an extra space in front of the first column name:

The second file, .pandas, shows the input after fixing this problem - indeed this enabled cleansumstats.sh pipeline to run without errors.

[pappewaio]

It seems you managed to solve the issue yourself, am I correct?

For the future, it is possible to access more log-files with more specific error reporting if you enter the work directory specified at the last line of your output above. But remember, the path you see is within the image, so if you want to access the workdir from outside the image it will be present in the tmp/fake-home/ directory, which is automatically mounted every time the image is being called. It might be a nice feature to make them more easily available directly in the user specified "outdir".

[ofrei]

Thanks!

Thanks for pointing to the tmp/fake-home folder - indeed, it had very useful info:

Yes, I solved this using python's pandas to read a .csv, then 'strip' all column names to get rid of spaces, and save back to disk. It's nice if cleansumstats.sh can do this, but it's not a showstopper - after all it's quite weird to have spaces in column names.

[pappewaio]

I think we can add a comment in the FAQ on how to do this removal of spaces in the beginning. Maybe it is even worth having a whole docs section only about how to handle strangeness in sumstats.