ttbbRun2Bamboo merge requestshttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests2023-03-16T20:02:48+01:00https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/91Split OOA and b-fragmentation2023-03-16T20:02:48+01:00Sebastien WertzSplit OOA and b-fragmentation- Split the signal and the OOA background into ttb and ttbb. This unfortunately requires editing the samples file (luckily, in a single place)
- Add the b-fragmentation systematic as a 1% simultaneous variation of all gen and reco level ...- Split the signal and the OOA background into ttb and ttbb. This unfortunately requires editing the samples file (luckily, in a single place)
- Add the b-fragmentation systematic as a 1% simultaneous variation of all gen and reco level jets with hadronFlavour=5. This required a small hack to bamboo itself, to be able to add variations to the reco-level jets on top of those already present from the JEC/JER. I'll make a PR for that soon.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/90produce pdf alphaS variations by default for full pdf samples2022-09-12T12:47:15+02:00Kyle Cormierproduce pdf alphaS variations by default for full pdf samplesThis will produce the alphaS variation templates (weights 101 and 102) for the 5FS ttbar samples which can then be transfered onto the 4FS ttB samples (which don't have the alphaS variations stored).
These will now be produced in the de...This will produce the alphaS variation templates (weights 101 and 102) for the 5FS ttbar samples which can then be transfered onto the 4FS ttB samples (which don't have the alphaS variations stored).
These will now be produced in the default 'simple' pdf scheme.
A little ugly since we assume the right weight positions, but I think thats fine for our purposes.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/89Add plotter to retrieve normalization effect of theory systematics at ttB level2022-08-17T15:02:23+02:00Sebastien WertzAdd plotter to retrieve normalization effect of theory systematics at ttB levelI figure it was easier to do it just once in a separate plotter, than to squeeze this into the unfoldingPlotter.
This produces a YAML file that looks like this:
```
powheg_4FS:
pdfDown: 0.9712332799989172
pdfUp: 1.02876745519370...I figure it was easier to do it just once in a separate plotter, than to squeeze this into the unfoldingPlotter.
This produces a YAML file that looks like this:
```
powheg_4FS:
pdfDown: 0.9712332799989172
pdfUp: 1.0287674551937036
psFSRDown: 1.0012873802848208
psFSRUp: 0.9990931462360679
psISRDown: 1.0018730077617866
psISRUp: 0.9985271701356928
qcdMuFDown: 1.03460182342414
qcdMuFUp: 0.9749808885020838
qcdMuRDown: 1.3418727639495327
qcdMuRFDown: 1.3417211362100763
qcdMuRFUp: 0.6918932791716198
qcdMuRUp: 0.721693279919291
topPt: 0.9838305594458754
powheg_5FS:
pdfDown: 0.9491977352020184
pdfUp: 1.0507965242645174
psFSRDown: 1.1519917536600055
psFSRUp: 0.8984246972509904
psISRDown: 1.0248103633581689
psISRUp: 0.9806493662957967
qcdMuFDown: 1.0290797453027525
qcdMuFUp: 0.9766607859089645
qcdMuRDown: 1.1257989795798686
qcdMuRFDown: 1.1532766512659771
qcdMuRFUp: 0.8655383354678515
qcdMuRUp: 0.8897398944103276
topPt: 0.9838065093955597
```
We'll now just need this in the unfolding code to add those as normalization uncertainties on ttB_OOA, while rescaling the shape variations to remove those normalization effects.Sebastien WertzSebastien Wertzhttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/88Fix bJet HT, adapt binning, uniformize electron eta cut, add top pt reweighting2022-08-16T18:26:31+02:00Sebastien WertzFix bJet HT, adapt binning, uniformize electron eta cut, add top pt reweightingWe were using the jet HT in the bJet HT observables. I've made an attempt at the binning, and made one run without systematics to adapt it further.We were using the jet HT in the bJet HT observables. I've made an attempt at the binning, and made one run without systematics to adapt it further.Sebastien WertzSebastien Wertzhttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/87Full JES split; flag for JER splitting; new colour scheme2022-07-28T13:57:21+02:00Sebastien WertzFull JES split; flag for JER splitting; new colour scheme- Add `--jes-scheme` argument to the plotters to select JEC splitting scheme (default is still `Merged`, other choices are `All` and `Total`). For the complete splitting, only the sources relevant for us have been activated.
- **Note...- Add `--jes-scheme` argument to the plotters to select JEC splitting scheme (default is still `Merged`, other choices are `All` and `Total`). For the complete splitting, only the sources relevant for us have been activated.
- **Note**: the `FlavorPureXXXX` variations are included, but will only be correct with additional changes to bamboo **and** the CMSJMECalculators package, for which I'll make a PR soon...
- Add `--no-split-jer` flag to use single JER variation
- Update colour scheme in config fileSebastien WertzSebastien Wertzhttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/86Add reco muon SFs2022-06-24T19:15:36+02:00Sebastien WertzAdd reco muon SFsTitle says all!Title says all!https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/85Add separate correlated muR+F variation, fix up/down directions2022-06-24T19:13:57+02:00Sebastien WertzAdd separate correlated muR+F variation, fix up/down directionsIt might be interesting to try to also include the correlated muR+F variation as a separate nuisance (or we might be asked to do it at some point). Even if don't, it's not much work to produce so we'll already have it in.
Also fixed the...It might be interesting to try to also include the correlated muR+F variation as a separate nuisance (or we might be asked to do it at some point). Even if don't, it's not much work to produce so we'll already have it in.
Also fixed the directions in the scale variations so that "down" also corresponds to scaling the weight down.
I've also added protections against non-existent theory weights for the alternative ttbb samples (in some cases the branch is there but only contains dummy weights).https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/84Binning adjustments; change gen jet to 25; add abseta(bb)2022-06-24T18:34:47+02:00Sebastien WertzBinning adjustments; change gen jet to 25; add abseta(bb)I've tried to adjust the binnings for the observables for which either the correlations in Asimov are large, or the unblinded fit simply didn't converge, or the last bin contained a lot of overflow. In the latter case, I've sometimes eve...I've tried to adjust the binnings for the observables for which either the correlations in Asimov are large, or the unblinded fit simply didn't converge, or the last bin contained a lot of overflow. In the latter case, I've sometimes even added an extra bin to increase the range.
Ideally we would have made several incremental runs to check all of these, and the gen jet pt cut change, but time is short so this is a bit of a bet that things will turn out OK...
The binning for abseta(bb) is a wild guess, so if it doesn't work we'll just skip it this time.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/83skimmer: changes for june 2022 synchronization round2022-06-24T19:11:02+02:00Sebastien Wertzskimmer: changes for june 2022 synchronization round- Update lumisections used for synchronization
- Store jet infos for first 4 jets
- Also store b-tag correction weight for each jet individually
The latter required to make some (transparent) changes to the functions returning the b-tag...- Update lumisections used for synchronization
- Store jet infos for first 4 jets
- Also store b-tag correction weight for each jet individually
The latter required to make some (transparent) changes to the functions returning the b-tagging corrections.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/82Workaround for 2016 btag mistag SFs2022-06-21T14:56:50+02:00Sebastien WertzWorkaround for 2016 btag mistag SFsAdd `--btag-fix` argument to use mistag SFs from 2016preVFP in postVFP.Add `--btag-fix` argument to use mistag SFs from 2016preVFP in postVFP.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/81Fix ttZ cross section2022-06-17T10:05:09+02:00Sebastien WertzFix ttZ cross sectionhttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/80Update ttV and ttH colours2022-06-16T09:37:48+02:00Sebastien WertzUpdate ttV and ttH coloursTo be in sync with https://gitlab.cern.ch/kcormier/uzh-ttbb-unfolding/-/blob/master/configs/full_analysis_dnn.ymlTo be in sync with https://gitlab.cern.ch/kcormier/uzh-ttbb-unfolding/-/blob/master/configs/full_analysis_dnn.ymlhttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/79Unfolding plotter: add argument to enable control regions, do not produce the...2022-06-14T09:15:22+02:00Sebastien WertzUnfolding plotter: add argument to enable control regions, do not produce them by defaultThis will make running things a little faster.This will make running things a little faster.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/78Add argument to decorrelate statistical b-tagging uncertainties between worki...2022-06-14T14:01:50+02:00Sebastien WertzAdd argument to decorrelate statistical b-tagging uncertainties between working points, fix b-tag SF evaluation bug, adjust control plotsArgument is `--decorr-btag` (by default nothing changes wrt before)Argument is `--decorr-btag` (by default nothing changes wrt before)https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/77Fix extra_jet_DR unfolding binning2022-06-07T09:23:14+02:00Sebastien WertzFix extra_jet_DR unfolding binninghttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/76More binning adjustments2022-06-05T11:22:51+02:00Sebastien WertzMore binning adjustmentsI've tried to make the finer bins showing larger fluctuatios a bit wider, at the expense of pushing up the upper edge of the distribution.
I have reduced the deltaPhi observable from 7 to 6 bins. I've hesitated to do the same for the |e...I've tried to make the finer bins showing larger fluctuatios a bit wider, at the expense of pushing up the upper edge of the distribution.
I have reduced the deltaPhi observable from 7 to 6 bins. I've hesitated to do the same for the |eta| distributions, but overall they look OK.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/75Fix unfolding selection names for ttbar processes2022-05-23T17:30:24+02:00Sebastien WertzFix unfolding selection names for ttbar processesMake sure the selections that are used to build the plots and the config files have a consistent name throughout.Make sure the selections that are used to build the plots and the config files have a consistent name throughout.https://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/74Changes for extra sample running2022-05-31T09:17:36+02:00Umberto MolinattiChanges for extra sample runninghttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/73Changes for extra sample running2022-05-20T17:43:31+02:00Umberto MolinattiChanges for extra sample runninghttps://gitlab.cern.ch/swertz/ttbbRun2Bamboo/-/merge_requests/72Fix order of PS weights for UL samples2022-05-18T15:53:13+02:00Sebastien WertzFix order of PS weights for UL samplesThere was a change in the Pythia version for UL that impacts the order of PS weights in TopNanoAOD: https://twiki.cern.ch/twiki/bin/viewauth/CMS/HowToPDF#Available_weights
So, in principle no difference for us, but I also noticed that t...There was a change in the Pythia version for UL that impacts the order of PS weights in TopNanoAOD: https://twiki.cern.ch/twiki/bin/viewauth/CMS/HowToPDF#Available_weights
So, in principle no difference for us, but I also noticed that the up and down direction were reversed (not that it makes any practical difference either).