Module grandma::plugins::distributions

Probability Distributions Plugins

This module containes plugins that simulate probability distributions on the nodes. It also has trackers used to see when queries and sequences are out of distribution.

Structs

BayesCategoricalTracker	Computes a frequentist KL divergence calculation on each node the sequence touches.
Categorical	Simple probability density function for where things go by count Stored as a flat vector in the order of the node addresses.
CategoricalTree	Zero sized type that can be passed around. Equivilant to ()
DiagGaussian	Node component, coded in such a way that it can be efficiently, recursively computed.
DiagGaussianTree	Zero sized type that can be passed around. Equivilant to ()
Dirichlet	Simple probability density function for where things go by count
DirichletBaseline	Trains a baseline by sampling randomly from the training set (used to create the tree) This baseline is not realistic.
DirichletTree	Zero sized type that can be passed around. Equivilant to ()
GrandmaCategorical	Zero sized type that can be passed around. Equivilant to ()
GrandmaDiagGaussian	Zero sized type that can be passed around. Equivilant to ()
GrandmaDirichlet	Zero sized type that can be passed around. Equivilant to ()
KLDivergenceStats	Tracks the non-zero (all KL divergences above 1e-10)

Traits

ContinousBayesianDistribution
ContinousDistribution
DiscreteBayesianDistribution
DiscreteBayesianSequenceTracker	Tracks the KL divergence for a given distribution.
DiscreteDistribution