uvex_transients.models.core.priors

Statistical prior distributions for uvex_transients.models.

This module defines the Prior interface and a collection of one-dimensional probability distributions used to describe model-parameter priors and generate random realizations.

Continuous priors are defined primarily through their log-probability density and support. From these, the base class provides probability-density, cumulative-distribution, and generic numerical-sampling functionality. Subclasses may override the sampling implementation when a more efficient analytic or library-backed generator is available.

The module also supports degenerate and discrete priors, for which probability mass functions and specialized sampling behavior replace the usual continuous density interface.

All priors operate on plain numerical coordinates. Physical units, parameter transformations, and other model-specific semantics are handled by the surrounding parameter infrastructure.

Classes

ConstantPrior(value)	Prior which always returns a single constant value.
DiscretePrior(values, probabilities)	Discrete weighted prior.
ExponentialPrior(scale)	Exponential prior.
LogNormalPrior(mean, sigma)	Log-normal prior.
MixturePrior(components, weights)	Weighted mixture of component priors.
NormalPrior(mean, sigma)	Gaussian prior.
PowerLawPrior(alpha, lower, upper)	Continuous power-law prior.
Prior()	Abstract base class for one-dimensional statistical priors.
TruncatedNormalPrior(mean, sigma, lower, upper)	Truncated normal prior.
UniformPrior(lower, upper)	Uniform prior over the interval [lower, upper).