LPC-DM / METXFit

0 stars 9 forks source link

b-tagging systematics #11

Closed mcremone closed 6 years ago

mcremone commented 6 years ago

1) For both resolved and monojet, b-tagging needs to be properly implemented to tie together the different btagging regions (resolved: pass, fail; monojet: 0,1,2 tags).

At the moment we have:

btag param 0.0 1 mistag param 0.0 1

We need to add something like:

zvv0tagsignal rateParam sig_0tag zvv (1.0@0@1) btag,mistag

where zvv is the process, 0tag is the btag bin, signal is the region. For example if we wanna get W+jets in the ttCR fail for resolved we will have:

wlnfailmuontop rateParam tmn_fail wjets (1.0@0@1) btag,mistag.

It is important you first understand why this is needed and why it has to be done this way.

2) A rate systematic on the heavy flavor fraction of the W/Z+jets process needs to be added for resolved and monojet. Not needed for boosted since it is already accounted for in the varying double-b efficiency. Take the mono-Hbb datacard as an example:

https://github.com/mcremone/MonoHiggsFit/blob/master/datacards/test_with_ttbar_fail.txt#L87-L88

@sudhaahuja @sundleeb @yguler @gurpinar @tiwariPC @SiewYan