Small fix to the cabinetry config, where the t-channel single top sample was accidentally using the s-channel input file. This has presumably existed from the beginning of the CMS ttbar setup, but has no further impact: all reference histogram files are still correct (they are produced prior to workspace creation).
Also adding a warning about the "data read" metric being buggy.
This also requires a new definition of pseudo-data to have sufficiently sensible fit results, which in turn requires a new set of reference histogram files. These will go in via the fix to #166.
Small fix to the cabinetry config, where the t-channel single top sample was accidentally using the s-channel input file. This has presumably existed from the beginning of the CMS ttbar setup, but has no further impact: all reference histogram files are still correct (they are produced prior to workspace creation).
Also adding a warning about the "data read" metric being buggy.
This also requires a new definition of pseudo-data to have sufficiently sensible fit results, which in turn requires a new set of reference histogram files. These will go in via the fix to #166.