pymc.HurdleGamma#

class pymc.HurdleGamma(name, psi, alpha=None, beta=None, mu=None, sigma=None, **kwargs)[source]#

Hurdle Gamma log-likelihood.

\[\begin{split}f(x \mid \psi, \alpha, \beta) = \left\{ \begin{array}{l} (1 - \psi) \ \text{if } x = 0 \\ \psi \frac{\text{GammaPDF}(x \mid \alpha, \beta))} {1 - \text{GammaCDF}(\epsilon \mid \alpha, \beta)} \ \text{if } x=1,2,3,\ldots \end{array} \right.\end{split}\]

where \(\epsilon\) is the machine precision.

Parameters:
psitensor_like of float

Expected proportion of Gamma draws (0 < psi < 1)

alphatensor_like of float, optional

Shape parameter (alpha > 0).

betatensor_like of float, optional

Rate parameter (beta > 0).

mutensor_like of float, optional

Alternative shape parameter (mu > 0).

sigmatensor_like of float, optional

Alternative scale parameter (sigma > 0).

Methods

HurdleGamma.dist(psi[, alpha, beta, mu, sigma])