Fitting Code does not converge properly

[uzzielperez]

While the templates are now being prepared properly, the fitting code still doesn't converge for 2017. It finishes for 2018 but the results are not expected. Need to figure out why this is happening.

The instructions for running the fitting code are here:

https://github.com/uzzielperez/diphoton-analysis/blob/ClosureTestCorrTemplates/FakeRateAnalysis/ClosureTestTemplateFits/README.md

One possible issue is that the ntuples have a finer binning and we might want to have a coarser binning for the closure test like before. We might need to find a way to merge some pt bins in the fit code.