bilby_pipe_process_mcmc
This executable will process an ongoing bilby_mcmc analyses to produce a combined result. The executable is assumed to be run from the submit directory (e.g., where the .ini file is) and on the submit machine.
Note
If your analysis uses HTCondor file transfer (transfer-files=True
), the
resume files required for processing will not exist on the submit machine until
the job has completed. You will need to use bilby_pipe_htcondor_sync
to first sync the files to the submit machine, then run
bilby_pipe_process_bilby_mcmc
.
usage: bilby_pipe_process_mcmc [-h] [-o OUTPUT_DIRECTORY] [--npool NPOOL]
[--thin-by-nact THIN_BY_NACT] [--plot]
[--convergence-check]
[--pt-rejection-sample PT_REJECTION_SAMPLE]
[-n NAMING_METHOD]
[--filter-match FILTER_MATCH]
[--filter-antimatch FILTER_ANTIMATCH]
[--no-conversion-function]
[--minimum-independent-samples MINIMUM_INDEPENDENT_SAMPLES]
directory
Positional Arguments
- directory
Directory to process: this should be the “outdir” specified in the configuration file
Named Arguments
- -o, --output-directory
Directory to write outputs, if not given directory is used.
- --npool
Multiprocess using npool
Default:
1
- --thin-by-nact
Overwrite the thin-by nact used.
- --plot
Create plots for each individual sampler.
Default:
False
- --convergence-check
Create convergence check plots.
Default:
False
- --pt-rejection-sample
If True, use pt_rejection_sampling.
Default:
False
- -n, --naming-method
Naming method for output. If ‘datetime’, name the file by thetime it was written. If ‘final’ name it as if it was final.
Default:
'datetime'
- --filter-match
Filter globbed files against this: anything not containing filter-match is discarded. Can use regex for matching.
- --filter-antimatch
Filter globbed files against this: anything containing filter-antimatch is discarded. Can use regex for matching.
- --no-conversion-function
If given, ignore the conversion_function
Default:
False
- --minimum-independent-samples
The minimum number of independent samples per chain.
Default:
50