_prepare_fit_setup¶

ell1fit.pipeline._prepare_fit_setup(parameters, requested_parameter_names, likelihood_func, parameters_with_unc, observation_length, model, template_funcs=None, weights=None, tolerance=1e-08, user_priors=None)[source]¶

Collect fit parameters, priors, factors, and initial fit values.

user_priors is a list of ell1fit.priors.PriorSpec overrides. It feeds both the priors and the scaling: see ell1fit.priors.user_prior_sigmas() for why a prior that does not also set the local scale would start the walkers outside itself.

Returns:

FitSetup: The bundle defining what is being fitted. Must be built after the templates exist: the per-file Phase_i offset they determine is what ell1fit.priors.assign_logpriors() centres that parameter’s prior on.