tfp.substrates.numpy.math.hpsd_logdet

Computes log|det(matrix)|, where matrix is a HPSD matrix.

Given matrix computes log|det(matrix)|, where matrix is Hermitian positive Semi-definite matrix.

matrix Floating-point Tensor of shape [..., N, N]. Represents a Hermitian positive semi-definite matrix.
cholesky_matrix (Optional) Floating-point Tensor of shape [..., N, N] that represents a Cholesky factor of matrix.

hpsd_logdet Scalar Tensor, retaining the batch shape of matrix.