This directory holds the skimmer & slimmer code (slimmer/monojet_TEMP.C) and the selection macro (makePlots.py). Its usage is briefly explained below.
The skimmer input is a nero file (preferablly from bambu). This code will drop the unnecessary branches, and add an other tree with some newly calculated information for the analysis. This will also replace some of the branches in the events tree depending which control region we would like to check such as met (it will become fakemet if you have 1 or 2 leptons).
In the future, this code can also be used to calculate shifted / smeared collection for systematic uncertainty calculations.
Note that, if the input file format changes, the slimmer and skimmer will have to change (at least the .h file) This is kind of annoying I agree.
At the current implementation monojet_TEMP.C is a TSelector. There is a small script written to interface the files to it properly. This script is: new_simpleRunner.csh (great naming I know). The usage is:
./new_simpleRunner.csh sampleNAME
This script creates a tchain using the files specified in the sampleNAME. This sampleNAME is a txt file in the /sourceFiles directory. It has list of all the files and their physical locations. The runner script will chain "events" tree of these files and run the monojet.C on them. It also will merge the "all tree" and write out at the same output file at the end. It further will save a histogram to be used in the plotting step for the effective number of events.
Now we are the step of the selection macro. The selection macro is the makePlots.py At the end of this script you can specify the variable you would like to draw and also the channel (signal or the control regions or even all). "selection.py" script will hold the different selection to be called for each channel. This script relies on a simple TDraw. The weights (lumi, mc weight, etc) are applied on the fly.
The input files to this plotter is given in LoadData.py There you can adjust the input file names, the colors, the xsec etc. Here we also decalre the other trees (such as the one created with more info) as a friend to our main events tree.
To run the selector and plotter do:
./makePlots.py -b -q -l
Make sure you have a directory called /test where the output histograms will be written to. This chain also produces the cut based data cards. There is a small annoyance to be fixed, which is when there are multiple bins for a process (such as Zvv) it creates multiple entries. Easy to fix will take care of it soon.
Once we have the signal samples we can test the sensitivity too.
The whole chain has been tested and the preliminary plots have been circulated. We are missing met filters which effects the overall distributions.