Prior distributions

Prior distributions encode prior beliefs about parameters in Bayesian inference. They represent knowledge or assumptions about parameter values before observing data.

Base classes

Parametric priors

Common continuous probability distributions for parameters.

Normal(mu, sigma[, name])	Normal (Gaussian) prior distribution.
Uniform([lower, upper, name])	Uniform prior distribution.
HalfNormal(*, sigma[, mu, name])	Half-normal prior distribution.
Gamma(alpha, beta[, mu, name])	Gamma prior distribution.
Beta(alpha, beta, lower, upper[, name])	Beta prior distribution for arbitrary [a, b] intervals.
LogNormal(mu, sigma[, name])	Log-normal prior distribution.
Cauchy(m, gamma[, name])	Cauchy prior distribution.
StudentT(ndof, mu, sigma[, name])	Student's t-distribution prior.
Exponential(lam[, name])	Exponential prior distribution.
Poisson(lam[, name])	Poisson prior distribution.

Non-parametric priors

Empirical priors constructed from data samples.

KDE(samples[, bandwidth, name])

Kernel Density Estimate prior from samples.