MixtureCandidate

class MixtureCandidate(val partition: DataPartition, components: List<ComponentCandidate>)(source)

One assembled mixture: a partition, one fitted component per group, and the mixing weights implied by the group sizes.

Parameters

partition

the partition the components were fitted to

components

the fitted components, one per group, in group order

Constructors

Link copied to clipboard
constructor(partition: DataPartition, components: List<ComponentCandidate>)

Types

Link copied to clipboard
object Companion

Properties

Link copied to clipboard

The fitted components, in group order.

Link copied to clipboard

The fitted component distributions, in group order.

Link copied to clipboard

The number of components.

Link copied to clipboard

The number of free parameters of the fitted mixture, as an information criterion should charge them: one fewer than the number of components for the mixing weights, since they are constrained to sum to one, plus every estimated parameter of every component including any estimated shift.

Link copied to clipboard
Link copied to clipboard

The mixing weights, taken as the empirical group proportions. These sum to one.

Functions

Link copied to clipboard

The observed-data log-likelihood of this candidate over the supplied observations, computed by log-sum-exp, together with the diagnostics that say whether it is comparable with another candidate's.

Link copied to clipboard

The assembled candidate as a continuous distribution, suitable for any scoring model, plot, or goodness-of-fit test that accepts one.

Link copied to clipboard
open override fun toString(): String

A readable description naming each component and its weight.