tf.contrib.distributions.quadrature_scheme_lognormal_gauss_hermite

Use Gauss-Hermite quadrature to form quadrature on positive-reals. (deprecated)

tf.contrib.distributions.quadrature_scheme_lognormal_gauss_hermite(
    loc, scale, quadrature_size, validate_args=False, name=None
)

Note: for a given quadrature_size, this method is generally less accurate than quadrature_scheme_lognormal_quantiles.

Args
loc	float-like (batch of) scalar Tensor; the location parameter of the LogNormal prior.
scale	float-like (batch of) scalar Tensor; the scale parameter of the LogNormal prior.
quadrature_size	Python int scalar representing the number of quadrature points.
validate_args	Python bool, default False. When True distribution parameters are checked for validity despite possibly degrading runtime performance. When False invalid inputs may silently render incorrect outputs.
name	Python str name prefixed to Ops created by this class.

Returns
grid	(Batch of) length-quadrature_size vectors representing the log_rate parameters of a Poisson.
probs	(Batch of) length-quadrature_size vectors representing the weight associate with each grid value.