Reference / binference.validation.ExactPosterior

binference.validation.ExactPosterior

class binference.validation.ExactPosterior(points, weights, log_density)

Posterior on quadrature nodes over the simplex (K <= 3).

Attributes:

points(n, K) array

Quadrature nodes (fractions).

weights(n,) array

Normalised posterior weights (quadrature weight times density), summing to one.

log_density(n,) array

Unnormalised log posterior density of the fractions at each node (its maximum locates the mode; the weights also carry the node spacing).

__init__(points, weights, log_density)

Methods

__init__(points, weights, log_density)
cov()
quantile(q, k)	Quantiles of the marginal posterior of fraction k.

Attributes

mean
std
points
weights
log_density

quantile(q, k)

Quantiles of the marginal posterior of fraction k.

Parameters:

qfloat or array_like

Probabilities in [0, 1].

kint

Bin index.