Updated files for migration to git (sync)
Improved iTTbar region plots; activated all plots for Appendix B; removed old way to get peak as a function of pileup; improved fitting for new way of getting pileup peak correction
Added function for going from local to global significance; switched ofsf plots back on in kin plot list; added old frank_t variable (called poor frank variable since it's the primitive way of getting it); completely revamped the of/sf plots with sys+stat uncertainty in the top plot and sys uncert in the ratio plot as a line along with stat directly in the points (like our money plot)
Added calculation of global significance
Feature-packed new version: - Allow to make mll plots after a given JZB cut for MC only (a kind of closure test) - do same for standard JZB cuts (0,50,80,100,150,200,250) - Adapted the legend for that kind of plots to include the JZB cut - At the moment these plots are disabled for data as I'm looking into MC closures but they can easily be re-enabled (look for 'Skipping data altogether') - Switched DY keyword back from M-20_CT10_ to DYJetsToLL - Reactivated LM point in Bpred plots (so one can nicely see the signal shape) - Completely reworked MC yields: - Automatically provides yields for 'residual MC samples' and compares them, including the differences - Skips irrelevant entries (flavor-symmetric ones) - Added and commented out a new way to get the response (with a different function, [0]+[1]/x + [2]/pow(x,2) ) - Leads to same result, but would overcomplicate JZB function when used in full glory, and does not lead to any visible improvement) - Added tau yields with errors for DY closure - Added lines denoting boundaries of mll regions in iTTbar check
Made several changes to iTTbar control region; added a bunch of plots to illustrate R(SF/OF) in iTTbar control region; added peak finding algorithm as a function of the number of vertices
Adapted JZB closure tests for new ttbar closure (with new ttbar samples); added several new plots for ttbar systematic tests; improved compatibility with different top pair production sample names
Added mll distribution after JZB cut in DY closure test; added quick test of peak position as a function of mll in testing
Able to produce OFSF plots for ETH and Aachen at the same time now
Added several nice variables for the signal region, such as tilde{S}_{T}, M_{T2} and JZB
Adapted MET notation for labels
Added ttbar tests, added illustration of variation of peak correction and response correction on ttbar estimate; added a bunch of SFOF plots
Finalized PU study
Updated systematics plots: now all samples are normalized to unit area and the shape is compared; to get a good picture of the actual systematic and not just statistical fluctuations we optimize the binning; to be able to still see trends we use 7 bins; the systematic is assumed to be symmetric about the central value, hence the variation is obtained solely based on up and down, not on the central value (as massivedecaybin sample is not recommended); also added illustration of PU dependence of width and peak position of JZB distribution
Removed top mass as source of systematic uncertainty
Updated API for picking up events
Added generator level plots (gen leptons and gen met)
Added all plots for Aachen as well
Added code to produce systematic plots much more quickly (just up/down ratio); should be used to get quick results, and the full blown function should be used to get all plots
Also updated 'save with ratio and sys band' function to not eat the supplied canvas; will commit updated Plotting_Functions soon
Updated ratio plots to new format (which doesn't each the tvirtualpad you pass it); added systematic plots (for scale and matching)
Added plots illustrating MC systematics
Added several plots illustrating the discussed set of systematic uncertainties on MC for the new version of the AN; also added the possibility to pass separate variables for MC and data (as is required e.g. for JZB)
Added function to compare data to mc including mc systematics (pu and jes) and draw them in ratio
Updated HT plots
Added HT plots
Updating THStack implementation
Added possibility to produce quick comparisons between two and three jet scenario, plus a 'plain' kinematic plotting function that doesn't assume any cuts
made data to data comparison (run comparison) clearer, with updated legends etc.
Updated plotting functions to use the new dileptonic ttbar sample for systematics; produce zjets closure for different mass regions (20<mll<70, Z window, mll>110)
box mll cut lowered to 70 in preparation of z-day
Adapted mll ranges (20<mll<70 and mll>120 instead of 15<mll<50 and 50<mll<70); adapted corresponding y ranges in histos
made drawing signal on kin plots optional; adapted y range for ttbar syst histo
Activated PURW for DY closure test (also made sure that if it's deactivated the peak position is harmonized); unified ratios for dy closure tests; fixed legend (DY with tau legend said DY->tau tau but it's DY-> ee,mm,tautau); fixed a whole bunch of conflicts with newly introduced kin plots that would overwrite old ones; adapted ttbar closure binning (going up to 400 with limited stats is not the way to go); adapted cuts for new plots;
Turned off underflow bin globally. Switched on overflow bin for plots requiring it. Removed superfluous pred/obs ratio function which is in the general prediction function anyway; adapted the range of the prediction & observation histos so the overflow bin is shown
Not switching under/overflow bin on after peak finding - this is now to be done individually in each function requiring it
One more plot.
Last fixes for "new" chunk of data. Added low-mass DY.
Added leading b-jet pt distribution.
Some fixes to the plots.
Lower mass cut at 15 GeV...
Fixes.
Updated peak finding to allow for different numbers of jets
Storing peak position; deactivated underflow/overflow bin while fitting
More plots in preparation for comparison and control regions.
Added protection against signal region. Prepared consistent list of plots. Also save ratio in SF/OF comparison
Made peak finding over/underflow resistant; adapted plots for high control region
Added sideband option for storage (required for scans)
Fixed accidentally overwritten section and two special sideband cases
Added option to switch off sidebands for classic JZB
Pimped up the OF/SF comparison plots.
Added some kinematic plots as requested, and ZJets estimate study
Added met plots for (JZB<0) and (JZB>0)
New kinematical plots added.
Restoring previous commit (1.40: response determination: use the Z peak)
Removed conflict for cJZB
Improved mll prediction plot.
Fix signal histogram in kinematic plots.
Fix name in JZB distributions. Make ratio in Bpred_MC only up to 2.0
- Implemented uncertainty on sigma of peak finder (cosmetics); - add background prediction on mll plot in signal region; - fine-tuning in kinematic plots; - added signal on kinematic and comparison plots; - implemented JES plot for OF; - more kinematic and comparison distributions; - response determination: use the Z peak, subtract OF; - optimized binning for Bpred ratio.
Removed sidebands for correction in computation of response correction for classic search (as we're not in the preselection with njets)
Adapted response computation: Now subtracting OF for inclusive search (and OF,OFSB, and SFSB for classic search); using MINOS for fitting
Removed shortcut (had deactivated some plots)
Updated Z+Jets systematic uncert plot: First doing a Z->ee,mm plot only (vetoing taus) with flipping, followed by ee,mm,tautau plot with all control regions
Added possibility to compare ee and mm jzb distributions
Several improvements: Deactivate PU for ttbar closure to improve statistics, added possibility to print out data yields in each bin, made mll plots after jzb cut non-logarithmic, deactivated under/overflow
Turned off over/underflow bin for some plots; loading uncertainty directly from PlottingSetup; added more intuitive binning for zjets
Added safety check to response correction
added option for global peak finding or individual peak finding; added alternative binning for ttbar closure test
Removed superfluous under/overflow protection; using eemm response correction for emu (instead of just average); writing PURW weight on screen; writing numbers for Z+Jets tests on screen;
using mc specific jzb variable for constructing corrected mc expression
Added separate peak finding for electrons and muons
Updated response correction to correct separately for electron and muon response
Adapted systematics drawn on closure/sensitivity/results plots, and activated it
Added custom binning for ZJets plot (because last bin contained almost no events); corrected bug in sideband assessment table (numbers were switched);
Adapted peak finding for btag case; removed unnecessary legends on bpred plot (referring to fit that isn't drawn anymore); adapted bpred ratio range (otherwise lm4 is outside the drawn range);
Added systematic error histo to prediction plot (but not drawing it yet)
Added new kinematic plots (for number of jets with OSOF), removed some that were removed at the JZBtree level (jet variables); now doing response with PURW
made last_configuration more easily readable
Using 3D reweighted events for ZJets systematics
included additional met plot
Added option to have customized binning for shape analysis; added btag option to response correction
adapted x scale for mll plot
fixed some uint vs int comparisons; commented out/removed unused variables; made Wall happy
forgot a brace in an if clause when porting changed. fixed.
Ported changes from paper to this version. In matters of style, changes were overruled
also doing diboson plots for main variable now
Added diboson plots (pred/obs and jzb plots)
Added new plot to show mll distribution as predicted and observed; expanded saved last configuration (for peak error in data)
ported paper changes to latest version
Added met vs jzb plot
Adapted y range of bpred ratio
Added option for signal vs bg plot to consider different signal than lm4
Added mll plot which essentially just plots the mll distribution for the different signal samples
Added masspeak restriction variable to last_config file output
Added check to see if all samples go down to the pt cut; this also serves to see which backgrounds increase/decrease when lowering/raising a pt cut
Added check for QCD in MC
made plot range more easily customizable
Initial commit of Ice Cream versions
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, select a symbolic revision name using the selection box, or choose 'Use Text Field' and enter a numeric revision.