em

Documentation

Overview

Package em is the groot implementation of the expectation-maximization algorithm for finding the most likely paths through the graphs Currently, there is no weighting for each path, as they are all so similar in length and the nodes have already been weighted. This may change after I've thought on it some more...

Index

• type EMrunner
  • func NewEM(nIterations int, mIterations int, paths map[uint32][]byte, ...) (*EMrunner, error)
  • func (EMrunner *EMrunner) Return() (int, []float64, error)
  • func (EMrunner *EMrunner) Run() error

Types

type EMrunner

type EMrunner struct {
	// contains filtered or unexported fields
}

EMrunner is the type to run the EM algorithm

func NewEM

func NewEM(nIterations int, mIterations int, paths map[uint32][]byte, lengths map[uint32]int, ecMap map[uint64][]uint32, counts map[uint64]float64) (*EMrunner, error)

NewEM is the constructor

func (*EMrunner) Return

func (EMrunner *EMrunner) Return() (int, []float64, error)

Return is the method to return EM results to the haplotype pipeline process

func (*EMrunner) Run

func (EMrunner *EMrunner) Run() error

Run is the method to run the EM algorithm