VHHAnalysis

VHH Analysis TOOLS

---

After set up CMS basic environment by cvmfs, one can set up the working environment by

source initial.sh

Other dependent packages can be setup with conda. Relavant envs has been listed in the LINK(under development...)

---

TOOLS

Analysis sample preparing

The Analysis sample preparing tool can be found in Tools, the slim_training_samples.py. Here's an example:

python slim_analysis_samples.py -m 16 -p 8 -d /PATH/TO/FILE/

Only specific branches will be kept after slimming, which can be set in analysis_branch.py.

Training sample preparing

Different script will be used to combine the Analysis samples to generate training samples.

Coupling Merge Tool

The coupling merge tool can be found in Tools, the CouplingMergingTool.py. Here's an example:

python CouplingMergingTool.py -m 4 -v 1 -n 6 -hh VHH -c c2v -r -30 30 -d 0 -w 10

-m is the number of threads to be used, -v is the verbose level, -n is the basic samples to be used to form the basic metrix, -c defines the coupling type(such as c2v,kl,cv,kt), -r sets the range of the scan, -d is the trigger if we do merge, -w is the new coupling of the merged sample, -hh defines the Di-Higgs Model (now we have updated models for GGF, VBF, VHH processes).

---

Training

BDT

XGBoost Toolkit will be used.

NN(??)

Consider Pytorch and Keras as candidates.