Module em

Source
Expand description

Expectation-maximisation (“EM”) algorithms for SFS inference.

Modules§

likelihood
Types describing likelihood of the data given an SFS.
stopping
Stopping rules used for deciding convergence for EM algorithms.

Structs§

Inspect
A combinator for types that allows inspection after each E-step.
StandardEm
A runner of the standard EM algorithm.
Window
A window of block SFS estimates, used in window EM.
WindowEm
A runner of the window EM algorithm.

Traits§

Em
A type capable of running an EM-like algorithm for SFS inference using data in-memory.
EmSite
A type of SAF site that can be used as input for EM.
EmStep
An EM-like type that runs in steps.
StreamEmSite
A type of SAF site that can be used as input for streaming EM.
StreamingEm
A type capable of running an EM-like algorithm for SFS inference by streaming through data.

Type Aliases§

ParallelStandardEm
A parallel runner of the standard EM algorithm.