Set up the Repo

To install this package:

cd $CMSSW_BASE/src/
cmsenv
git clone https://github.com/Andrew42/EFTGenReader.git EFTGenReader
scram b

Note: In the wrapperEFTGenPlots.py file, make sure the GEN_PLOTS_DIR variable points to the directory that you want the output directory to be created.

Run the code

If reading files from the grid (e.g. for central samples) make sure to initialize your voms proxy first

voms-proxy-init -voms cms -valid 192:00

To run the code:

cd $CMSSW_BASE/src/EFTGenReader/GenReader/test
cmsRun EFTGenReader_cfg.py json_sample_name
python wrapperEFTGenPlots.py --web-dir=/absolute/or/relative/path/to/some/output/directory --dir=sub-directory_name fpath1 fpath2 etc...

Where json_sample_name is the name of an entry in the datasets.json file. The non-DAS samples are only reachable from earth. fpath1, fpath2, etc. are the (relative) file paths to the root files produced by running EFTGenReader_cfg. Each root file passed to wrapperEFTGenPlots will have the same named histograms overlayed with one another on a single plot.

cmsRun Options

The EFTGenReader_cfg supports a number of command-line options for configuring the cmsRun job, which are implemented using the CMSSW VarParsing package:

Command-line Options:
    dataset=Name
        Name of the dataset as it appears in the JSON file
    test=True
        Indicates that the output root file should use the dummy TEST name
    debug=True
        Run in debug mode
    normType
        How to normalize the histograms; 0 - no norm, 1 - unit norm (default), 2 - xsec norm
    intgLumi
        Value to scale all histograms by (no effect for unit norm mode)
Examples:
    # basic running
    cmsRun EFTGenReader_cfg.py dataset=central_ttH
    # run in test mode
    cmsRun EFTGenReader_cfg.py test=True dataset=central_ttH
    # perform no histogram normalizations
    cmsRun EFTGenReader_cfg.py normType=0 dataset=central_ttH

chainGenReader

The python script chainGenReader.py can be used to do multiple cmsRun commands in series. All of the config options available to EFTGenReader_cfg can be specified at the command-line for chainGenReader, which will then be passed on to the underlying cmsRun call. All samples in the chain will be ran with the same command-line options. Type python chainGenReader.py --help for a complete list of available options.

Misc. Other

The information used to find the samples is stored in the datasets.json file located in GenReader/data/JSON. Currently, the code is only able to run on samples that are located locally on hadoop, or on DAS. Files are found by looking in the directory specified by the loc entry. DAS files are found by querying via the command line utility dasgoclient, which should be available from in any recent CMSSW release.

Additional datasets can be added/modified/removed from the datasets.json file either by hand (not recommended) or by making use of the update_datasets.py script. The script shows some examples of how to add/modify/remove entries from the dataset.json file via the DatasetHelper class. Currently, there is no way to edit the JSON file entirely from the command-line.

Note: The versions of the TH1EFT and WCFit classes in this repo have been modified relative to the ones in the ttH-Multilepton repository. Here we have re-enabled the storing of the fit errors as well as switched back to the alternative TH1EFT::Scale member function.