phitter.fit.likelihood#

Module Contents#

Classes#

log_likelihood_chisq

log_likelihood_chisq is an object to obtain chi squared log likelihood. Instantiate this object with an observables object to indicate the observed observables.

log_likelihood_chisq_weighted_obs_type

log_likelihood_chisq_weighted_obs_type is an object to obtain chi squared log likelihood, while weighing different observation types equally. This weighting scheme may be useful for cases where number of a given observation type is much larger than that of another.

log_likelihood_chisq_weighted_filts_and_obs_type

log_likelihood_chisq_weighted_filts_and_obs_type is an object to obtain chi squared log likelihood, while weighing different observation types, and filters within each observation type equally. This weighting scheme may be useful for cases where number of a given observation type or filt is much larger than that of another.

API#

class phitter.fit.likelihood.log_likelihood_chisq(observations, *args, **kwargs)#

Bases: object

log_likelihood_chisq is an object to obtain chi squared log likelihood. Instantiate this object with an observables object to indicate the observed observables.

observationsobservables

observables object, specified with the observations to be fit.

Initialization

evaluate(model_observables)#
class phitter.fit.likelihood.log_likelihood_chisq_weighted_obs_type(observations, *args, **kwargs)#

Bases: phitter.fit.likelihood.log_likelihood_chisq

log_likelihood_chisq_weighted_obs_type is an object to obtain chi squared log likelihood, while weighing different observation types equally. This weighting scheme may be useful for cases where number of a given observation type is much larger than that of another.

The output log likelihood is calculated in the following way: log_like_total = (log_like_t1)/(n_t1) + (log_like_t2)/(n_t2) + … (see e.g., Lam+ 2022)

Instantiate this object with an observables object to indicate the observed observables.

observationsobservables

observables object, specified with the observations to be fit.

Initialization

evaluate(model_observables)#
class phitter.fit.likelihood.log_likelihood_chisq_weighted_filts_and_obs_type(observations, *args, **kwargs)#

Bases: phitter.fit.likelihood.log_likelihood_chisq

log_likelihood_chisq_weighted_filts_and_obs_type is an object to obtain chi squared log likelihood, while weighing different observation types, and filters within each observation type equally. This weighting scheme may be useful for cases where number of a given observation type or filt is much larger than that of another.

The output log likelihood is calculated in the following way: log_like_total = (log_like_t1)/(n_t1) + (log_like_t2)/(n_t2) + … (see e.g., Lam+ 2022)

Instantiate this object with an observables object to indicate the observed observables.

observationsobservables

observables object, specified with the observations to be fit.

Initialization

evaluate(model_observables)#