TrackingDstar

Lambda Merged Analysis

Identify and split merged hits in the CMS Pixel Tracker

Analyzer + Notebook setup instruction

Setting up the Analyzer

• Setup: (do not run any extra command from the moment you connect to fermilab/lxplus6, not lxplus-slc7)

  • cd /uscms_data/d3/hichemb/princeton/project1/ (go to your home directory).
  • source /cvmfs/cms.cern.ch/cmsset_default.csh (or sh for bash).
  • scram list (gives you list of CMSSW and check compatibility with your current architecture).
  • echo $SCRAM_ARCH (outputs: 'slc6_amd64_gcc476' for me, not compatible with the CMSSW environment I want to use (10_4_0), so I change architecture).

• If using CSH: Setenv SCRAM_ARCH slc6_amd64_gcc700 . If using BASH: SCRAM_ARCH=newarchictecture

• export SCRAM_ARCH

  • echo $SCRAM_ARCH (should output: 'slc6_amd64_gcc700').
  • cmsrel CMSSW_10_4_0 (I chose this one).
  • cd CMSSW_10_4_0/src
  • cmsenv
  • (DO NOT MAKE ANY EXTRA DIRECTORY).
  • git clone https://github.com/nhaubrich/TrackingDstar
  • scram b -j 8 (should get a few warnings).
  • cd TrackingDstar/LambaAnalyzer/test/

Running/Making root files

Run locally (fastest, good for testing):

• cmsRun local_trkeffanalyzer_MC_GeneralTracks_cfg.py (runs on files from SingleNeutrino dataset I copied over)

To run crab jobs:

• voms-proxy-init -voms cms (use certificate, apply for grid certificate if you don't have one)
• source /cvmfs/cms.cern.ch/crab3/crab.csh OR .sh (depending on what you use)
• crab submit crab_MC_truth_cfg.py (can change RECO file used by editing crab_MC_truth_cfg.py)
• (once you submit crab use 'crab status…(copy command from there)' to check on status.)

The following are broken/outdated:

• trkeffanalyzer_data_GeneralTracks_cfg.py may be broken (missing input data ?))
• cmsRun trkeffanalyzer_Data_GeneralTracks_cfg.py (outdated)
• cmsRun trkeffanalyzer_MC_GeneralTracks_cfg.py (outdated)

ROOT Post-Processing

After running analyzer

• hadd output1.root *.root (add root files if you have multiple files). (for crab jobs: hadd the files in your eos directory by simply including the path everytime before filename).
• cp /eos/uscms/store/user/hichemb/RelValMinBias_13/crab_lambdaanalysis_relval11/190715_135818/0000/output1.root . To copy file from eos to local dir.
• python makeLambdaHists.py output1.root output2.root [remove duplicate lambdas (i.e. set their mass to -999) , this script can be merged with the initial python file).
• python addBranchesForNNTraining.py output2.root (add pixel branches, outputs ofile2.root, only first lambda with mass>0 and keeping just important stuff) [UPDATE: now you can run this script as follow: python addBranchesForNNTraining.py output2.root output3.root rather than edit the name of the output in the file everytime].
• python converRootToPandas.py ofile2.root (turns root file pixel info into pandas df, outputs pixelTrain.h5, can view panda file using python shell and command pandas.read_hdf("pixelTrain.h5")). [after the changes I made, run it as follow: python converRootToPandas.py output3.root output3.h5]

Jupyter Notebooks

• Set up CERNBOX and SWAN.
• Import Github URL: https://github.com/nhaubrich/MergedHits, now you have all the need script and files.
• Go to SWAN bash terminal and run the following command pip install --user tables.
• Connect to your lxplus account and go to directory cd /eos/user/h/hboucham/ (where 'h' is your username first letter and 'hboucham' is your username, change accordingly).
• Make a file called startup.sh using your favorite text editor (Vim obviously) and include these 2 lines:
  • #!/bin/bash
  • export PYTHONPATH=$CERNBOX_HOME/.local/lib/python2.7/site-packages:$PYTHONPATH
• When you start SWAN in the configuration setting copy the following in environment $CERNBOX_HOME/startup.sh and use Python 2 ('95a') [ very important !] and SLC7 (this option might be outdated).
• Run do_merged.ipynb, if you did everything above it should run with no errors. If you get an error prompting you to import PyTables then you messed up somewhere.
• Enjoy your plots.

Monte Carlo Generation

Instructions for generating $\Lambda$ gun events using pythia.

First, Setup CMSSW and add the MC package

1. Login to fnal with your username. $ kinit [email protected] $ ssh [email protected]
2. Open a bash shell. $ bash
3. Go to your d3. $ cd /uscms_data/d3/bbonham
4. Make a new directory (e.g. LambdaMC) for the MC and switch to it. $ mkdir LambdaMC $ cd LambdaMC
5. Get a CMSSW release and activate the environment. $ cmsrel CMSSW_10_5_0 $ cd CMSSW_10_5_0/src $ cmsenv
6. Add the MC components to CMSSW. $ git cms-addpkg Configuration/Generator
7. Compile. $ scram b -j 4

Next, Prepare and Run a MC Config File

8. Go to the dir with example config files. $ cd Configuration/Generator/python
9. Copy a config file to use as a template and give it a name (e.g. myLambdaGun_cfi.py). $ cp SingleElectronPt35_pythia8_cfi.py
10. Edit the config file to suit your needs. $ vim myLambdaGun_cfi.py
11. Generate a RAW .root file and a RECO .root file with 10 events. You'll may need to alter the options below to suit your needs. $ cmsDriver.py myLambdaGun_cfi.py -n 10 --mc --step GEN,SIM,DIGI,L1,DIGI2RAW,RAW2DIGI,RECO --conditions auto:mc --eventcontent RAWSIM,RECOSIM --datatier RAW,RECO

Submitting crab jobs for MC generation

12. Go to the dir with example config files. $ cd Configuration/Generator/python
13. Run teh crab config file to submit 10 jobs with 100 events each, you can edit BOTH config files to adjust the number fo jobs and events: $crab submit crab_config.py
14. Check for job status: $crab status -d crab_projects/crab_crab1000
15. Or resubmit failed jobs: $crab resubmit -d crab_projects/crab_crab1000
16. Your jobs' output should be located in your eos directory (/eos/uscms/store/user/hichemb/MinBias/ for example)
17. Don't forget to change the name of the the next jobs in the config file or deleting crab_projects/crab_crab1000.