Understand scale variation changes via correctionlib implementation

[alexander-held]

#119 changed the pt_scale_up variations, in particular the following changes:

sum(all_histograms[:, "4j2b", "ttbar", "pt_scale_up"].values())

Prior to this PR, the propagation of the pT changes on the observable itself was not taken into account, so the bin-by-bin distribution in this region is expected to differ regardless before and after the PR. It is currently unclear though why also the total event count would change in this region.

The 4j1b region does not seem to be affected by this.

cc @Andrew42

[ekauffma]

As mentioned above, originally, pT variations were not propagated through when calculating the observable in region "4j2b". This is fixed with the correctionlib implementation.

Thus different events are cut off by the histogram binning. sum(all_histograms[:, "4j2b", "ttbar", "pt_scale_up"].values()) does not include overflow and underflow events, hence the difference. This problem also affected the "pt_res_up" variation, although the discrepancy was masked by the fact that the values differ randomly with each run.

Changing the histogram binning to include all events which were cut off reveals that the total event count remains the same.

[alexander-held]

Thanks a lot for tracking this down @ekauffma!

[alexander-held]

A backport for v1 is done via #163.