bayespy.utils.random.gamma_entropy

bayespy.utils.random.gamma_entropy(a, log_b, gammaln_a, psi_a, a_psi_a)[source]

Entropy of \mathcal{G}(a,b).

If you want to get the gradient, just let each parameter be a gradient of that term.

Parameters:
a : ndarray

a

log_b : ndarray

\log(b)

gammaln_a : ndarray

\log\Gamma(a)

psi_a : ndarray

\psi(a)

a_psi_a : ndarray

a\psi(a)