search

elaird / supy

analyze events stored in TTrees in parallel
8 stars 7 forks source link

Tweaks #161

Closed betchart closed 11 years ago

betchart commented 11 years ago

Hi Ted and Davide,

Here are a few tweaks I would like included in supy.

Burt

gerbaudo commented 11 years ago

Hi Burt,

Thanks for these tweaks. They all look very reasonable to me. I tried the code with a minimal analysis, https://github.com/davidegerbaudo/supy-ttbar/blob/test-ll-fit/analyses/compare_ll_kin.py with one single input file, I see that analysisLooper.printStats is not being called (it even looks like analysisLooper.mergeFunc is not being called). The results in the pdf, however, are identical.

This behavior seems to be due to this change: https://github.com/elaird/supy/pull/161/files#L0L118 If this is expected, that's fine, please merge :-) (just reporting what see) Thanks.

Here is an example output (between the two tries, I made sure to change one histogram in the analysis file so that supy is forced to loop again): http://pastebin.com/HPgeD0qp

betchart commented 11 years ago

Hi Davide,

Thanks for testing. That is not intended behavior. My tests had been on the batch queue, and the printStats still fires when pulling together the batch results, so I hadn't noticed it that it stopped working for local looping. I did reproduce your observation, so I will investigate.

Burt

gerbaudo commented 11 years ago

Hi Burt, Thanks for the last commit. Just a quick note to confirm that it works as expected. Feel free to merge.