cogwheel.posterior.submit_likelihood_maximization

cogwheel.posterior.submit_likelihood_maximization(eventname, mchirp_guess, approximant, prior_name, parentdir, scheduler='slurm', n_hours_limit=2, scheduler_cmds=(), overwrite=False, **kwargs)

Submit a job that runs main(), which maximizes the likelihood for an event. This will initialize Posterior.from_event() and Posterior.refine_reference_waveform and save the Posterior to JSON in the appropriate eventdir (per Posterior.get_eventdir).

Parameters:

eventnamestr

Event name, e.g. ‘GW150914’.

mchirp_guessfloat

Approximate chirp mass (Msun).

approximantstr

Approximant name.

prior_namestr

Key in gw_prior.prior_registry.

parentdiros.PathLike

Path to top directory where to save output.

scheduler{‘slurm’, ‘lsf’}

n_hours_limitint

Hours until jobs are terminated.

sbatch_cmdssequence of strings

Commands for the scheduler. E.g., (’–mem-per-cpu=4G’,) for slurm, or (‘-R “span[hosts=1] rusage[mem=4096]”’) for lsf.

overwritebool

Whether to overwrite preexisting files. False (default) raises an error if the file exists.

**kwargs

Optional keyword arguments to Posterior.from_event(), must be JSON-serializable.