0. Documentation

• mkShapesRDF
• ROOT RDataFrame class reference
• HTCondor batch system
• Condor commands
• PyRoot tutorials
• SWAN
• Monte Carlo particle numbering scheme
• The Review of Particle Physics (PDG)

1. Installation

Log in to the CERN LXPLUS cluster.

ssh -Y -l <username> lxplus.cern.ch -o ServerAliveInterval=240

Switch to a Bash shell.

bash -l

Clone the project.

git clone https://github.com/BlancoFS/mkShapesRDF.git
cd mkShapesRDF
git checkout Run3

Once you have cloned the project you need to modify install.sh.

emacs -nw install.sh

Add the following line right before python -m pip install -e ".[docs,dev,processor]".

unset SSH_ASKPASS

Now you can proceed with the installation.

./install.sh

Get the analysis configuration.

git clone https://github.com/piedraj/HEP

2. Always do

Everytime you start a new session you need to follow these steps.

ssh -Y -l <username> lxplus.cern.ch -o ServerAliveInterval=240
bash -l
cd mkShapesRDF
source start.sh

Currently the analysis can be performed on v7 or v9 samples. To choose one set simply go to the corresponding folder.

cd HEP/Full2018_v7/
cd HEP/Full2018_v9/

3. Produce the analysis histograms

Action	Command
Compile	mkShapesRDF --compile 1
Run on local	mkShapesRDF --operationMode 0 --folder . --doBatch 0 --limitEvents 10
Run on batch	mkShapesRDF --operationMode 0 --folder . --doBatch 1
Check filled jobs	mkShapesRDF --operationMode 1 --folder .
Resubmit jobs	mkShapesRDF --operationMode 1 --folder . --resubmit 1
Merge root files (batch only)	mkShapesRDF --operationMode 2 --folder .
Available arguments	mkShapesRDF --help

Always proceed as follows:

1. Modify the code. This is where the analaysis development starts.
2. Compile. If you miss this step the implemented changes won't be considered when running the code.
3. Run on local. This is a very important step, as it allows you to check if any error has been introduced with the latest modifications.
4. Run on batch.
5. Merge root files.

4. Check job status

condor_q

If you need to cancel all the submitted jobs.

condor_rm <username>

5. Plot the analysis histograms

mkPlot --showIntegralLegend 1 --onlyPlot cratio --fileFormats png

If needed, the available arguments can be printed.

mkPlot --help

When running on batch check that you have produced these v7 or v9 default distributions.

6. How to know a tree content

Both data and Monte Carlo (MC) files have the event variables stored in a ROOT TTree object named Events. Once you have followed the instructions below, you just need to open the file TreeContent.h to see all the variables.

root -l /eos/cms/store/group/phys_higgs/cmshww/amassiro/HWWNano/Summer20UL18_106x_nAODv9_Full2018v9/MCl1loose2018v9__MCCorr2018v9NoJERInHorn__l2tightOR2018v9/nanoLatino_TTJets__part52.root
Events->MakeClass("TreeContent")
.q

7. Share on the web

Once the analysis plots have been produced, the best way to look at them is by creating a webEOS site.