bregman.application.distribution package

Subpackages

• bregman.application.distribution.exponential_family package
  • Subpackages
    • bregman.application.distribution.exponential_family.categorical package
      • Submodules
      • bregman.application.distribution.exponential_family.categorical.categorical module
      • Module contents
    • bregman.application.distribution.exponential_family.gaussian package
      • Submodules
      • bregman.application.distribution.exponential_family.gaussian.dissimilarity module
      • bregman.application.distribution.exponential_family.gaussian.gaussian module
      • bregman.application.distribution.exponential_family.gaussian.geodesic module
      • Module contents
    • bregman.application.distribution.exponential_family.multinomial package
      • Submodules
      • bregman.application.distribution.exponential_family.multinomial.dissimilarity module
      • bregman.application.distribution.exponential_family.multinomial.geodesic module
      • bregman.application.distribution.exponential_family.multinomial.multinomial module
      • Module contents
  • Submodules
  • bregman.application.distribution.exponential_family.exp_family module
    • ExponentialFamilyDistribution
      • ExponentialFamilyDistribution.F()
      • ExponentialFamilyDistribution.k()
      • ExponentialFamilyDistribution.pdf()
      • ExponentialFamilyDistribution.t()
    • ExponentialFamilyManifold
      • ExponentialFamilyManifold.kl_divergence()
      • ExponentialFamilyManifold.t()
  • Module contents
• bregman.application.distribution.mixture package
  • Subpackages
    • bregman.application.distribution.mixture.discrete_mixture package
      • Submodules
      • bregman.application.distribution.mixture.discrete_mixture.discrete_mixture module
      • Module contents
    • bregman.application.distribution.mixture.ef_mixture package
      • Submodules
      • bregman.application.distribution.mixture.ef_mixture.ef_mixture module
      • Module contents
  • Submodules
  • bregman.application.distribution.mixture.mixture module
    • MixingDimensionMissMatch
    • MixtureDistribution
      • MixtureDistribution.weights
      • MixtureDistribution.distributions
      • MixtureDistribution.pdf()
    • MixtureManifold
      • MixtureManifold.distributions
      • MixtureManifold.distribution_to_point()
      • MixtureManifold.jensen_shannon_divergence()
      • MixtureManifold.kl_divergence()
      • MixtureManifold.point_to_distribution()
      • MixtureManifold.set_distributions()
    • MixturePoint
      • MixturePoint.display()
  • Module contents

Submodules

bregman.application.distribution.distribution module

class bregman.application.distribution.distribution.DistributionManifold(theta_generator: Generator, eta_generator: Generator | None, display_factory_class: type[MyDisplayPoint], dimension: int)

Bases: ApplicationManifold[MyDisplayPoint], Generic[MyDisplayPoint, MyDistribution], ABC

Abstract class for statistical manifolds of distributions.

abstract distribution_to_point(distribution: MyDistribution) → MyDisplayPoint

Converts a distribution object to a point in the manifold.

Parameters:

distribution – Distribution object to be converted.

Returns:

Point corresponding to the distribution object.

jensen_shannon_divergence(point_1: Point, point_2: Point) → ndarray

Jensen-Shannon-Divergence between two points (their distributions) in the manifold.

Parameters:

• point_1 – Left-sided argument of the KL-Divergence.

• point_2 – Right-sided argument of the KL-Divergence.

Returns:

Jensen-Shannon-Divergence between point_1 and point_2’s corresponding distributions.

kl_divergence(point_1: Point, point_2: Point) → ndarray

KL-Divergence between two points (their distributions) in the manifold.

Parameters:

• point_1 – Left-sided argument of the KL-Divergence.

• point_2 – Right-sided argument of the KL-Divergence.

Returns:

KL-Divergence between point_1 and point_2’s corresponding distributions.

abstract point_to_distribution(point: MyDisplayPoint) → MyDistribution

Converts a point to a distribution object.

Parameters:

point – Point to be converted.

Returns:

Distribution object corresponding to the point.

Module contents