BugFix for btagSF variation
The variation of the btagSF were filled in the wrong way because of the c-flavour masks.
The light flavour variations were filled with 1 instead of the nominal btagSF when the cferr*
variations were computed.
This was causing one-sided uncertainty bands for cferr variations.
N.B.: Implemented the btagSF central value for JES variations.