The Initial State Analysis (ISA) framework uses ntuples produced with the FinalStateAnalysis framework. This framework constructs user defined initial states from final state objects and stores interesting variables in an output ntuple for further selection.

ISA requires HiggsCombine tool to produce limits and ROOT.

cmsrel CMSSW_7_4_9
cd CMSSW_7_4_9/src
cmsenv
git cms-init
cd recipe
./recipe.sh
cd $CMSSW_BASE/src
scram b -j 16

The primary analyzer is accessed via the run.py command. This command has paths to ntuples stored for convenient access. For example, to run the TT channel of the WZ analysis over all MC samples:

# Usage: run.py [analysis] [channel] [period] samples (unix wildcards allowed)
run.py WZ WZ 13 W* T* DY* Z*

Jobs can be submitted to the cluster using the --submit option:

./run.py --submit --jobName=testSubmit Hpp3l Hpp3l 13 D* T* W* Z* 

Plotting can be accomplished via the mkplots.py command:

# Usage: mkplots.py [analysis] [channel] [period] [options]
mkplots.py Hpp3l Hpp3l 13

Limits can be run via mklimits.py. This produces datacards able to be read by the HiggsAnalysis/CombinedLimit module.

The mklimits.py script can produce limits using three different methods: a purely MC driven method that estimates background from MC samples, a data-driven method with a user defined sideband and signal region, and a fakerate method (requires the fakerate option on the ntuple production, TODO).

# Usage: mklimits [analysis] [region] [period] [options]
mklimits.py Hpp3l Hpp3l 13

The datacards can then be processed with the processdatacards.py script:

# Usage: processdatacards.py [analysis] [region] [period] [options]
processdatacards.py Hpp3l Hpp3l 13

And finally, the limits can be plotted with plotlimits.py:

# Usage: plotlimits.py [analysis] [region] [period] [options]
plotlimits.py Hpp3l Hpp3l 13 -bp ee100