fccanalysis-run - run FCC analysis
fccanalysis run [-h | --help] [-i
INPUT_FILE [INPUT_FILE ...], --input INPUT_FILE
[INPUT_FILE ...]] [--files-list INPUT_FILE
[INPUT_FILE ...]] [-f LIST_FILE,
--input-file-list LIST_FILE] [--output OUTPUT]
[--nevents NEVENTS] [--test] [--bench]
[-j NCPUS | --ncpus NCPUS | --n-threads
NTHREADS] [-g] [--graph-path GRAPH_PATH]
[--use-data-source] [--output-dir OUTPUT_DIR]
[-a ANALYSIS_NAME | --analysis-name
ANALYSIS_NAME] [--apply-filepath-rewrites |
--no-filepath-rewrites] [-s SAMPLE_NAME |
--sample-name SAMPLE_NAME] [--n-chunks N_CHUNKS]
[--stride STRIDE] [--test-file TEST_FILE]
analysis-script -- [ANALYSIS ARGUMENTS]
fccanalysis-run will run analysis provided in the analysis
script on the current machine, i.e. locally. The analysis itself can be
divided into several stages, if desired. For all those stages
fccanalysis-run is used.
When using fccanalysis-run the analysis is running in the
managed mode, where the RDataFrame is steered by the framework and users can
control only some aspects of the running with additional global attributes,
see fccanalysis-script(7).
- analysis-script
- Path to analysis script.
- -h, --help
- Prints short help message and exits.
- -i INPUT_FILE
[INPUT_FILE ...], --input INPUT_FILE [INPUT_FILE
...]
- Directly specify input ROOT files for the analysis. Takes precedence over
the files provided in a file list or the files determined from the
process_list of the analysis.
- --files-list
INPUT_FILE [INPUT_FILE ...]
- [DEPRECATED] Directly specify input ROOT files for the analysis.
- --input-file-list
LIST_FILE
- Specify file containing list of input ROOT files to be used for the
analysis. Takes precedence over the files determined from the
process_list of the analysis. Each input ROOT file needs to be on a
separate line. Lines starting with "#" are ignored.
- --output
OUTPUT
- Specify output file name, otherwise dataset name used.
- --nevents
NEVENTS, --n-events NEVENTS
- Specify max number of events to process.
- --test
- Run over the test file.
- --use-data-source
- Load events through podio::DataSource.
- --output-dir
OUTPUT_DIR
- Set global output directory for the analysis.
- -a ANALYSIS_NAME,
--analysis-name ANALYSIS_NAME
- Set analysis name, mostly used for the output path.
- --apply-filepath-rewrites,
--no-filepath-rewrites
- Apply or disable path rewrites for files stored on shared file
systems.
- -s SAMPLE_NAME,
--sample-name SAMPLE_NAME
- Set sample name, mostly used for the output path. Used when running over
an independent sample (i.e. without a sample list).
- --n-chunks
N_CHUNKS
- Split the output into N_CHUNKS files. Used only when running over a
test file or an independent sample.
- --stride
STRIDE
- Specify how to stride through the events. Used only when running over a
test file or an independent sample.
- --test-file TEST_FILE
- Specify the test file to use when running with --test.
- --bench
- Output benchmark results to a JSON file.
- -j NCPUS,
--ncpus NCPUS (deprecated), --n-threads
NTHREADS
- Set number of jobs (threads) to use. The default number of threads to use
is 1. To use all threads available on the machine specify -1.
- -g, --graph
- The computational graph of the analysis will be generated.
- --graph-path
GRAPH_PATH
- Location where the computational graph of the analysis should be stored.
Only paths with .dot and .png extensions are accepted.
- ANALYSIS
ARGUMENTS
- Command line arguments which will not be parsed by the framework and will
be passed directly to the analysis script.
- FCCDICTSDIR
- Controls search path for the process dictionaries. The default value is
/cvmfs/fcc.cern.ch/FCCDicts/.
fccanalysis(1), fccanalysis-script(7)
There are many contributors to the FCCAnalyses framework, but the
principal authors are:
Clement Helsens
Valentin Volkl
Gerardo Ganis
Part of the FCCAnalyses framework.