CLsHypo

class hep_spt.CLsHypo(pf)

Bases: object

Represent an hypothesis to be used in the CLs method. The class is built from a given probability function.

Parameters

pf (scipy.stats.rv_frozen) – probability density/mass function.

Variables

func – probability density/mass function representing the hypothesis.

See also

CLsHypo_discrete, CLsHypo_continuous

Methods Summary

__call__(v)	Calculate the global probability for the given input value(s).
median()	Calculate the median of the distribution associated to this hypothesis.
percentil(prob)	Calculate the percentil associated to the given probability.

Methods Documentation

__call__(v)

Calculate the global probability for the given input value(s). If the arguments of the probability function are arrays, then it is assumed that “v” is either an array of length equal to the number of values per argument, or that it is an array of arrays following the latter structure. If “v” does not match any of the previous requirements, then the global probability is calculated assuming this value for all the sub-functions. Assuming “pmf” as “poisson.pmf”:

>>> import numpy as np
>>> from scipy.stats import poisson
>>> ha = cls_hypo(poisson, 4)
>>> ha(4) # Same as pmf(4, 4)
0.19536681481316454
>>> hb = cls_hypo(poisson, np.array([4, 8]))
>>> hb([4, 8]) # Same as pmf(4, 4)*pmf(4, 8)
0.02727057613800409
>>> hb([[4, 8], [6, 10]]) # Same as [pmf(4, 4)*pmf(8, 8), pmf(6, 4)*pmf(10, 8)]
>>> hb(4) # Same as pmf(4, 4)*pmf(4, 4)
0.011185197244103258

Parameters

v (numpy.ndarray) – input value(s).

Returns

Global probability.

Return type

float

median()

Calculate the median of the distribution associated to this hypothesis. This is equivalent to call CLsHypo.percentil(0.5).

Returns

Median of the distribution.

Return type

numpy.ndarray

percentil(prob)

Calculate the percentil associated to the given probability.

Parameters

prob (float) – probability.

Returns

Percentil.

Return type

numpy.ndarray