hepstats.utils#

hepstats.utils.fit#

Module for testing a fitting library validity with hepstats.

A fitting library should provide six basic objects:

model / probability density function

parameters of the models

data

loss / likelihood function

minimizer

fitresult (optional)

A function for each object is defined in this module, all should return True to work with hepstats.

The zfit API is currently the standard fitting API in hepstats.

`is_valid_parameter`(object)	Checks if a parameter has the following attributes/methods:
`is_valid_data`(object)	Checks if the data object has the following attributes/methods:
`is_valid_pdf`(object)	Checks if the pdf object has the following attributes/methods:
`is_valid_loss`(object)	Checks if the loss object has the following attributes/methods:
`is_valid_fitresult`(object)	Checks if the fit result object has the following attributes/methods:
`is_valid_minimizer`(object)	Checks if the minimzer object has the following attributes/methods:

`get_value`(value)
`eval_pdf`(model, x[, params, allow_extended])	Compute pdf of model at a given point x and for given parameters values
`pll`(minimizer, loss, pois[, init])	Compute minimum profile likelihood for fixed given parameters values.
`array2dataset`(dataset_cls, obs, array[, weights])	dataset_cls: only used to get the class in which array/weights will be converted.
`get_nevents`(dataset)	Returns the number of events in the dataset

`base_sampler`(models, nevents)	Creates samplers from models.
`base_sample`(samplers, ntoys[, parameter, ...])	Samples from samplers.