eccentricity_summary

ell1fit.eccentricity.eccentricity_summary(eps1, eps2, upper_limit_level=0.95, detection_sigma=3.0, flat_in_e_prior=False)

Summarize the eccentricity posterior, as a value or as an upper limit.

Parameters:

eps1, eps2array-like

Paired posterior samples, in physical units. See load_eps_samples() for getting them out of a finished run.

upper_limit_levelfloat

Credible level of the upper limit. Default 0.95.

detection_sigmafloat

Equivalent-Gaussian significance (see zero_eccentricity_exclusion()) the exclusion of \(e = 0\) has to reach before a median and error bars are quoted instead of an upper limit. Default 3.

flat_in_e_priorbool

Reweight the samples by \(1/e\) to report the answer under a prior flat in \(e\), undoing the \(p(e) \propto e\) implied by the flat priors on EPS1 and EPS2. The detection test is always made on the samples as drawn – it asks what the data say about the origin, which is not a question the radial prior should be allowed to move.

Returns:

dict

ECC_<percentile> for each of PERCENTILES; ECC_upper_limit and ECC_upper_limit_level (the limit is nan when the eccentricity is detected, since a limit is then not the thing to quote); ECC_upper_limit_3sigma and its _level, which unlike the above are reported either way; ECC_detected; ECC_zero_credibility and ECC_significance_sigma from zero_eccentricity_exclusion(); the OM_deg_* fields of _circular_summary(); every field of eps_component_summary(), under this same prior; ECC_nsamples; ECC_prior; and ECC_summary, the one-line form to paste into a paper draft.