jax.scipy.stats.dirichlet.logpdf#

jax.scipy.stats.dirichlet.logpdf(x, alpha)[source]#

Dirichlet log probability distribution function.

JAX implementation of scipy.stats.dirichlet logpdf.

The Dirichlet probability density function is

\[f(\mathbf{x}) = \frac{1}{B(\mathbf{\alpha})} \prod_{i=1}^K x_i^{\alpha_i - 1}\]

where \(B(\mathbf{\alpha})\) is the beta() function in a \(K\)-dimensional vector space.

Parameters:

x (Array | ndarray | bool_ | number | bool | int | float | complex) – arraylike, value at which to evaluate the PDF
alpha (Array | ndarray | bool_ | number | bool | int | float | complex) – arraylike, distribution shape parameter

Returns:

array of logpdf values.

Return type:

Array

jax.scipy.stats.dirichlet.logpdf