The histograms getting saved to disk now are no longer re-binned from the coffea output, so they're 25 bins from 50 to 550 GeV (20 GeV bin width). During template collection in cabinetry the lowed bin edge gets set to 110 GeV and the bins are merged 2->1, which previously was already done to the histograms saved to the root file.
This addresses the second item in #162 via an intermediate solution that can be replaced once https://github.com/scikit-hep/cabinetry/issues/412 is addressed. It makes use of https://cabinetry.readthedocs.io/en/latest/advanced.html#overrides-for-template-building to subsequently re-bin via
cabinetry
, which ends up less complicated that I had initially expected.The histograms getting saved to disk now are no longer re-binned from the coffea output, so they're 25 bins from 50 to 550 GeV (20 GeV bin width). During template collection in
cabinetry
the lowed bin edge gets set to 110 GeV and the bins are merged 2->1, which previously was already done to the histograms saved to the root file.