gcfit.probabilities.priors.UniformPrior

class gcfit.probabilities.priors.UniformPrior(param, edges, *, transform=False)

Flat uniform prior function.

Represents a normalized uniform prior likelihood distribution, defined by N bounding pairs which must all overlap smoothly. The returned likelihood value for any point between the minimum and maximum bounds is defined by the size of the distribution, and normalized to 1.

Bounds are evaluated on each call, in order to support dependant parameters, and the inv_value will be returned in the case of bound-pairs which do not overlap.

Parameters:

paramstr

Name of the corresponding parameter.

edgeslist of 2-tuple

List of all bound pairs (lower, upper). Bounds can be either a float, for a fixed bound, or a string name of another parameter, for a dependant bound.

transformbool, optional

Whether this is a prior likelihood or a prior transform function. Changes whether the distribution PDF or PPF is evaluated upon calling.

Attributes:

boundslist of 2-tuple

The bounds, as defined by the edges parameter.

See also

scipy.stats.uniform

Distribution class used for pdf/ppf evaluation.

Methods

__init__(param, edges, *[, transform])

Attributes

dependants
inv_mssg
inv_value
param