cophycollapse

package module
v0.0.0-...-8a86f7c Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: May 20, 2019 License: GPL-2.0 Imports: 14 Imported by: 0

README

cophycollapse

Documentation

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

func AssertNumChars 

func AssertNumChars(chars1, chars2 []float64)

AssertNumChars will check to make sure that two character vectors are the same lengths (as they always should be)

func AssertNumMis 

func AssertNumMis(chars1, chars2 []bool, node *Node)

AssertNumMis will make sure that the boolian slices are the same length

func AssertUnrootedTree 

func AssertUnrootedTree(tree *Node)

AssertUnrootedTree is a quick check to make sure the tree passed is unrooted

func BMParallelPruneRooted 

func BMParallelPruneRooted(n *Node, lab int)

BMParallelPruneRooted will prune BM branch lens and PICs down to a rooted node root node should be a real (ie. bifurcating) root added Aug 2018. will need later.

func BMPruneRooted 

func BMPruneRooted(n *Node)

BMPruneRooted will prune BM branch lens and PICs down to a rooted node root node should be a real (ie. bifurcating) root

func BMPruneRootedSingle 

func BMPruneRootedSingle(n *Node, i int)

BMPruneRootedSingle will prune BM branch lens and calculate PIC of a single trait down to a rooted node root node should be a real (ie. bifurcating) root

func BMPruneRootedSub 

func BMPruneRootedSub(n *Node, sites []int)

BMPruneRootedSub will prune BM branch lens and PICs down to a rooted node root node should be a real (ie. bifurcating) root

func CalcExpectedTraits 

func CalcExpectedTraits(tree *Node)

CalcExpectedTraits will plug in the expected values for missing traits under BM using the pruning/PIC ancestral state estimation approach

func CalcExpectedTraitsSub 

func CalcExpectedTraitsSub(tree *Node, sites []int)

CalcExpectedTraitsSub will plug in the expected values for missing traits under BM using the pruning/PIC ancestral state estimation approach

func CalcRootedLogLike 

func CalcRootedLogLike(n *Node, nlikes *float64, startFresh bool)

CalcRootedLogLike will return the BM likelihood of a tree assuming that no data are missing from the tips.

func CalcSiteMeans 

func CalcSiteMeans(nodes []*Node) (siteSum []float64)

CalcSiteMeans will calculate the mean value for all the sites in the matrix for which the site is not missing

func CalcTreeAIC 

func CalcTreeAIC(tree *Node, crit int) (aic float64)

func CalcUnrootedLogLike 

func CalcUnrootedLogLike(tree *Node, startFresh bool) (chll float64)

CalcUnrootedLogLike will calculate the log-likelihood of an unrooted tree, while assuming that no sites have missing data.

func ClusterMissingTraitsEM 

func ClusterMissingTraitsEM(tree *Node, cluster *Cluster, niter int)

ClusterMissingTraitsEM will iteratively calculate the ML branch lengths for a particular topology and cluster when doing the greedy site clustering procedure.

func ColumnMatrixToSlice 

func ColumnMatrixToSlice(m *mat.Dense) []float64

ColumnMatrixToSlice will convert a Dx1 matrix to a slice

func DMtoPhylip 

func DMtoPhylip(dm map[string][]float64, nodes []*Node) (lines []string)

func DirichletBranchLengthLogPrior 

func DirichletBranchLengthLogPrior(nodes []*Node, pr *BranchLengthPrior) float64

DirichletBranchLengthLogPrior will return the log prior probability of all branch lengths in a trejke

func EBExponentialBranchLengthLogPrior 

func EBExponentialBranchLengthLogPrior(nodels []*Node, means []float64) (prob float64)

EBExponentialBranchLengthLogPrior will give the exponential probability of each branch length when exponential mean == ML branch length

func EBNormalBranchLengthLogPrior 

func EBNormalBranchLengthLogPrior(nodels []*Node, means []float64) (prob float64)

EBNormalBranchLengthLogPrior will calculate the prior probability of a branch length from a normal distribution centered on the ML branch length estimate

func ExponentialBranchLengthLogPrior 

func ExponentialBranchLengthLogPrior(nodels []*Node, beta float64) (prob float64)

ExponentialBranchLengthLogPrior will give the exponential prior probability of observing a particular set of branch lengths on a tree, given shape lambda

func GIWStartingSampleMean 

func GIWStartingSampleMean(traits map[string][]float64) (mu0 *mat.Dense)

GIWStartingSampleMean will calculate the starting mean value as just the raw sample mean from all traits

func GammaTreeLengthPrior 

func GammaTreeLengthPrior(nodels []*Node, alpha, beta float64) float64

GammaTreeLengthPrior will calculate the prior probability of the current tree length

func GreedyIterateLengthsMissing 

func GreedyIterateLengthsMissing(tree *Node, sites []int, niter int)

GreedyIterateLengthsMissing will iteratively calculate the ML branch lengths for a particular topology and cluster when doing the greedy site clustering procedure.

func InitMissingValues 

func InitMissingValues(tree []*Node)

InitMissingValues will find the missing sites in a data matrix and plug in values corresponding to the mean of the remaining sites

func InitParallelPRNLEN 

func InitParallelPRNLEN(nodes []*Node)

InitParallelPRNLEN will set up empty slices for the prnlens

func IterateBMLengths 

func IterateBMLengths(tree *Node, niter int)

IterateBMLengths will iteratively calculate the ML branch lengths for a particular topology, assuming that traits are fully sampled at the tips. should use MissingTraitsEM if there are missing sites.

func IterateLengthsWeighted 

func IterateLengthsWeighted(tree *Node, cluster *Cluster, niter int)

IterateLengthsWeighted will iteratively calculate the ML branch lengths for a particular topology and cluster when doing the greedy site clustering procedure.

func LogGammaFn 

func LogGammaFn(x float64) float64

LogGammaFn will return the log of the gamma function of x. X should be provided as a float64, but be a whole number

func MakeMissingDataSlice 

func MakeMissingDataSlice(t *Node)

MakeMissingDataSlice will intialize the MIS attribute of node t by identifying traits with LARGE (ie. missing) values and updating MIS accordingly.

func MakeMissingMeansTip 

func MakeMissingMeansTip(n *Node, means []float64)

MakeMissingMeansTip will replace missing values with the mean across all tips for a single tip

func MakeRandomStartingBranchLengths 

func MakeRandomStartingBranchLengths(tree *Node)

MakeRandomStartingBranchLengths will initialize a tree with a set of random branch lengths

func MapContinuous 

func MapContinuous(t *Node, traits map[string][]float64, ntraits int)

MapContinuous maps the traits contained with in a (golang) map to the tips of a tree and initializes slices of the same length for the internal nodes

func MarkAll 

func MarkAll(nodes []*Node)

MarkAll will mark all of the nodes in a tree ARRAY

func Max 

func Max(l map[int][]int) (biggest int)

Max returns the maximum value in a map of ints used like a set

func MaxClustLab 

func MaxClustLab(l map[int]*Cluster) (biggest int)

MaxClustLabF64 returns the maximum value in a map of ints used like a set

func MaxClustLabBig 

func MaxClustLabBig(l map[int]*big.Float) (biggest int)

MaxClustLab returns the maximum value in a map of ints used like a set

func MaxClustLabF64 

func MaxClustLabF64(l map[int]float64) (biggest int)

MaxClustLabF64 returns the maximum value in a map of ints used like a set

func Mean 

func Mean(v []float64) float64

Mean wil calculate the mean from a slice of floats

func MeanInt 

func MeanInt(v []int) int

MeanInt wil calculate the mean from a slice of ints

func MissingRootedLogLike 

func MissingRootedLogLike(n *Node, startFresh bool) (sitelikes float64)

MissingRootedLogLike will return the BM log-likelihood of a tree for a single site, pruning tips that have missing data

func MissingTraitsEM 

func MissingTraitsEM(tree *Node, niter int)

MissingTraitsEM will iteratively calculate the ML branch lengths for a particular topology

func MissingUnrootedLogLike 

func MissingUnrootedLogLike(tree *Node, startFresh bool) (sitelikes float64)

MissingUnrootedLogLike will calculate the log-likelihood of an unrooted tree. It is deprecated and is only a convienience for testing LL calculation.

func PhyloVCV 

func PhyloVCV(nodes []*Node) *mat.Dense

PhyloVCV will return the variance-covariance matrix calculated from a tree with branch lengths

func PruneToStar 

func PruneToStar(tree *Node)

PruneToStar will prune brlens and traits to a root

func ReadContinuous 

func ReadContinuous(traitfl string) (map[string][]float64, int, int)

ReadContinuous will read in a PHYLIP-formatted trait file and return a map

func ReadFossils 

func ReadFossils(path string) (fos []string)

ReadFossils will read in a list of fossil tips one line at a time into a slice TODO: get this working

func ReadLine 

func ReadLine(path string) (ln []string)

ReadLine is like the Python readline() and readlines()

func ReadMCLoutput 

func ReadMCLoutput(clfl string) (clusters map[int]*Cluster)

func Rexp 

func Rexp(lambda float64) (e float64)

Rexp will draw a random exponential number

func SetIdentityMatrix 

func SetIdentityMatrix(dim int) *mat.Dense

SetIdentityMatrix will return an identiy matrix with dimensions ntax,ntax

func SingleSiteLL 

func SingleSiteLL(tree *Node, site int) (sitelike float64)

SingleSiteLL will return the likelihood of a single site

func SitewiseLogLike 

func SitewiseLogLike(tree *Node) (sitelikes []float64)

SitewiseLogLike will calculate the log-likelihood of an unrooted tree, while assuming that some sites have missing data. This can be used to calculate the likelihoods of trees that have complete trait sampling, but it will be slower than CalcRootedLogLike.

func SubDM 

func SubDM(nodes []*Node, clus *Cluster) map[string][]float64

SubDM will return the variance-covariance matrix calculated from a tree with branch lengths

func SubUnrootedLogLikeParallel 

func SubUnrootedLogLikeParallel(tree *Node, sites []int, workers int) (sitelikes float64)

SubUnrootedLogLikeParallel will calculate the log-likelihood of an unrooted tree, while assuming that some sites have missing data. This can be used to calculate the likelihoods of trees that have complete trait sampling, but it will be slower than CalcRootedLogLike.

func SymDenseConvert 

func SymDenseConvert(m *mat.Dense) *mat.SymDense

SymDenseConvert will convert a matrix of type *Dense to *SymDense

func TreeLength 

func TreeLength(nodes []*Node) float64

TreeLength will return the total length of a slice of nodes

func TritomyML 

func TritomyML(tree *Node)

TritomyML will calculate the MLEs for the branch lengths of a tifurcating 3-taxon tree

func TritomySubML 

func TritomySubML(tree *Node, sites []int)

TritomySubML will calculate the MLEs for the branch lengths of a tifurcating 3-taxon tree using only the sites indicated in sites

func TritomyWeightedML 

func TritomyWeightedML(tree *Node, weights map[int]float64)

TritomyWeightedML will calculate the MLEs for the branch lengths of a tifurcating 3-taxon tree

func WeightedUnrootedLogLike 

func WeightedUnrootedLogLike(tree *Node, startFresh bool, weights []float64) (sitelikes float64)

WeightedUnrootedLogLike will calculate the log-likelihood of an unrooted tree, while assuming that some sites have missing data. This can be used to calculate the likelihoods of trees that have complete trait sampling, but it will be slower than CalcRootedLogLike.

func WeightedUnrootedLogLikeParallel 

func WeightedUnrootedLogLikeParallel(tree *Node, startFresh bool, weights []float64, workers int) (sitelikes float64)

WeightedUnrootedLogLikeParallel will calculate the log-likelihood of an unrooted tree, while assuming that some sites have missing data. This can be used to calculate the likelihoods of trees that have complete trait sampling, but it will be slower than CalcRootedLogLike.

Types

type BranchLengthPrior 

type BranchLengthPrior struct {
	TYPE      string
	ALPHA     float64
	BETA      float64
	NTIPS     int
	SGAMMA    float64
	SFACT     float64
	CUR       float64
	LAST      float64
	CLUSTCUR  map[int]float64
	CLUSTLAST map[int]float64
	TREELEN   float64
	GAMMADIST distuv.Gamma
}

BranchLengthPrior is a struct for specifying and calculating different priors

func InitializePrior 

func InitializePrior(priorType string, nodes []*Node) *BranchLengthPrior

InitializePrior will set up a new instance of a prior

func (*BranchLengthPrior) Calc 

func (pr *BranchLengthPrior) Calc(nodes []*Node) float64

Calc will return the log prior probability of the branch length prior

func (*BranchLengthPrior) DrawDirichletBranchLengths 

func (pr *BranchLengthPrior) DrawDirichletBranchLengths(nodes []*Node) (lengths []float64)

DrawDirichletBranchLengths will randomly assign branchlengths from the compound dirichlet distribution

type CharAlignment 

type CharAlignment struct {
	MatSites     map[int]*mat.Dense // site index is the key, traits for each taxon are the values
	MatSitesT    map[int]mat.Matrix // same as above, but the site trait vectors are transposed
	MatSitesProd map[int]*mat.Dense // stores A*A^T for each site A
	SiteOrder    []string           // stores the order that traits are stored in for the untransposed trait vectors
	Dim          int
	NSites       int
}

CharAlignment will store all characters for a collapsed Gibbs clustering analysis, along with their transposed vectors

func InitTraitMatrices 

func InitTraitMatrices(traits map[string][]float64) *CharAlignment

InitTraitMatrices will initialize the CharAlignment struct

type Cluster 

type Cluster struct {
	Sites         []int // this stores all of the sites with a preference for this cluster
	BranchLengths []float64
	LogLike       float64
	SiteWeights   map[int]float64 // this will store the probability that each site in the MATRIX belongs here.
}

func (*Cluster) CalcLL 

func (c *Cluster) CalcLL(tree *Node)

func (*Cluster) WriteClusterPhylip 

func (c *Cluster) WriteClusterPhylip(nodes []*Node) string

type DPPGibbs 

type DPPGibbs struct {
	Aln             *CharAlignment
	Clusters        map[int]*MVNLike
	SiteAssignments map[int]int
	Prior           *GIWPrior
	Gen             int
	PrintFreq       int
	WriteFreq       int
	Threads         int
	Workers         int
	Alpha           float64
}

DPPGibbs is a struct to store the attributes of a collapsed DPP gibbs sampler for the BM phylo gene expression data

func InitGibbs 

func InitGibbs(traits map[string][]float64, prior *GIWPrior, gen, print, write, threads, workers int, alpha float64) *DPPGibbs

InitGibbs will initialize the parameters for the collapsed gibbs sampler.

func (*DPPGibbs) CalcMn 

func (chain *DPPGibbs) CalcMn(clust *MVNLike, N, kn float64, newSiteData *mat.Dense) *mat.Dense

CalcMn will calculate the mean vector of the posterior for a single cluster after a new site has been added

func (*DPPGibbs) CalcSn 

func (chain *DPPGibbs) CalcSn(clust *MVNLike, site int, N float64) (*mat.Dense, *mat.Dense)

CalcSn will calculate the VCV of the posterior distribution

func (*DPPGibbs) Run 

func (chain *DPPGibbs) Run()

Run will run MCMC simulations

type DistanceMatrix 

type DistanceMatrix struct {
	MatSites map[int][]float64 // pairwise comparison is the key, value is a slice containing the comparison between taxa across all sites
	//MatSitesT map[int]mat.Matrix // same as above, but the site trait vectors are transposed
	//MatSitesProd map[int]*mat.Dense // stores A*A^T for each site A
	PWDistLabs []int // stores the order that traits are stored in for the untransposed trait vectors
	//Dim       int
	NSites    float64
	IntNSites int
}

DistanceMatrix will store a continuous character matrix converted into a pairwise distance matrix for each site

func DM 

func DM(nodes []*Node) *DistanceMatrix

DM will return the variance-covariance matrix calculated from a tree with branch lengths

type GIWPost 

type GIWPost struct {
	MuN *mat.Dense // parameters are same as prior
	KN  float64
	S   *mat.Dense
	VN  float64
}

GIWPost stores values for the posterior distribution calculated analytically from GIWPrior and the phylo BM likelihood

type GIWPrior 

type GIWPrior struct {
	Mu0          *mat.Dense // prior on multivariate normal mean (ancestral state at root of tree)
	K0           float64    // belief in Mu0
	S0           *mat.Dense // prior on multivariate normal VCV
	V0           float64    // belief in S0
	PostPredProd *mat.Dense // this will store Mu0*Mu0^T*K0 (for use in calculating the posterior parameters)
	Mu0K0        *mat.Dense
	PPDensity    float64
}

GIWPrior is an object for the Gausian inverse Wishart prior used as conjugate prior for brownain phylogenetic VCV likelihood

func InitVCVPrior 

func InitVCVPrior(mu0 *mat.Dense, k0 float64, s0 *mat.Dense, v0 float64) *GIWPrior

InitVCVPrior will initialize the prior on the VCV matrix and BM mean

type HCSearch 

type HCSearch struct {
	Tree            *Node
	PreorderNodes   []*Node
	Clusters        map[int]*Cluster
	SiteAssignments map[int]int
	Gen             int
	Threads         int
	Workers         int
	RunName         string
	LogOutFile      string
	K               int
	PrintFreq       int
	CurrentAIC      float64
	NumTraits       float64
	Criterion       int
	SavedConfig     []*SiteConfiguration
	CurBestAIC      float64
	JoinLikes       map[int]map[int]float64
	SplitGen        int
	Alpha           float64
	NumPoints       float64
	ExpandPenalty   float64
	MinK            int
}

func InitEMSearch 

func InitEMSearch(tree *Node, gen int, k int, pr int, alpha float64) *HCSearch

func InitGreedyHC 

func InitGreedyHC(tree *Node, gen int, pr int, crit int, rstart bool, k int, runName string, splitgen int, alpha float64, minK int) *HCSearch

func TransferGreedyHC 

func TransferGreedyHC(tree *Node, gen int, pr int, crit int, clus map[int]*Cluster, siteAssign map[int]int, runName string, splitgen int, alpha float64) *HCSearch

func (*HCSearch) CalcRelLikes 

func (s *HCSearch) CalcRelLikes() (denom float64)

func (*HCSearch) CheckCluster 

func (s *HCSearch) CheckCluster(checkConfig *SiteConfiguration) (keep bool)

func (*HCSearch) ClusterString 

func (s *HCSearch) ClusterString() string

ClusterString will return a string of the current set of clusters

func (*HCSearch) NewSiteConfig 

func (s *HCSearch) NewSiteConfig() *SiteConfiguration

func (*HCSearch) PerturbedRun 

func (s *HCSearch) PerturbedRun()

func (*HCSearch) RefineSavedClusterings 

func (s *HCSearch) RefineSavedClusterings()

func (*HCSearch) RunEM 

func (s *HCSearch) RunEM()
type HCSearch struct {
	Tree            *Node
	PreorderNodes   []*Node
	Clusters        map[int]*Cluster
	SiteAssignments map[int]int
	Gen             int
	Threads         int
	Workers         int
	ClustOutFile    string
	LogOutFile      string
	K               int
	PrintFreq       int
}

func (*HCSearch) RunSingleHC 

func (s *HCSearch) RunSingleHC()

func (*HCSearch) SplitEM 

func (s *HCSearch) SplitEM()

func (*HCSearch) WriteBestClusters 

func (s *HCSearch) WriteBestClusters()

func (*HCSearch) WriteClusterTrees 

func (s *HCSearch) WriteClusterTrees()

type MVNLike 

type MVNLike struct {
	Sites []int
	//MeanData *mat.Dense
	DataSum *mat.Dense
	ProdSum *mat.Dense // gives the sum of A*A^T across all sites A in the cluster
	// contains filtered or unexported fields
}

MVNLike will store all data and params for each cluster

func (*MVNLike) ClusterSampleMean 

func (mvn *MVNLike) ClusterSampleMean(aln *CharAlignment) *mat.Dense

ClusterSampleMean will calculate the sample mean of all the sites in a particular cluster

type NGPost 

type NGPost struct {
	MuN    float64 // parameters are same as prior
	KN     float64
	AlphaN float64
	BetaN  float64
}

NGPost stores values for the posterior distribution calculated analytically from GIWPrior and the phylo BM likelihood

type Node 

type Node struct {
	PAR    *Node
	CHLD   []*Node
	NAME   string
	LEN    float64
	PRNLEN float64
	CONTRT []float64
	//VISITED bool
	MRK        bool
	MIS        []bool //this is a slice of bools that indicates whether the index is missing from CONTRT
	LL         []float64
	CONPRNLEN  []float64
	CLUSTLEN   map[int]float64
	PARTPRNLEN map[int]float64 // [][]float64
}

Node is a node struct to represent rooted and unrooted phylogenetic trees

func InternalNodeSlice 

func InternalNodeSlice(nodes []*Node) (inNodes []*Node)

InternalNodeSlice will return a slice containing only internal nodes

func MRCA 

func MRCA(n1, n2 *Node) *Node

MRCA will return the MRCA of two tips

func ReadTree 

func ReadTree(ts string) (root *Node)

ReadTree will parse a newick string into a node struct

func (*Node) AddChild 

func (n *Node) AddChild(c *Node) error

AddChild will add a child to a node

func (*Node) MarkAll 

func (n *Node) MarkAll()

MarkAll will unmark all of the nodes on a tree

func (*Node) NNodes 

func (n *Node) NNodes(count *int)

NNodes is a helper method that will return the number of internal nodes descending from n (including n)

func (Node) Newick 

func (n Node) Newick(bl bool) (ret string)

Newick will return a newick string representation of the tree

func (*Node) PostorderArray 

func (n *Node) PostorderArray() (ret []*Node)

PostorderArray will return an array of all the nodes in the tree in Postorder

func (*Node) PreorderArray 

func (n *Node) PreorderArray() (ret []*Node)

PreorderArray will return a preordered array of all the nodes in a tree

func (*Node) RemoveChild 

func (n *Node) RemoveChild(c *Node)

RemoveChild will remove a chidl from the slice of children associated with a node

func (*Node) Reroot 

func (n *Node) Reroot(oldroot *Node) *Node

Reroot reroots all the nodes represented in a graph on n

func (*Node) UnmarkAll 

func (n *Node) UnmarkAll()

UnmarkAll will unmark all of the nodes on a tree

func (*Node) UnmarkToRoot 

func (n *Node) UnmarkToRoot(oldroot *Node)

UnmarkToRoot will mark all of the nodes betwen the current node and the root to be recalculated

type NormalGammaPrior 

type NormalGammaPrior struct {
	Mu0    float64 // hyperparameter for expected mean pairwise distance
	K0     float64 // hyperparameter on the precision (1/variance) of the prior mean
	Alpha0 float64 // shape hyperpameter on the variance of the distance estimates
	Beta0  float64 // scale hyperparemeter on the variance of the distance estimates
	//PostPredProd float64 // this will store Mu0*Mu0^T*K0 (for use in calculating the posterior parameters)
	//Mu0K0        *mat.Dense
	PPDensity *distuv.StudentsT
}

NormalGammaPrior is an object for the Gausian inverse Wishart prior used as conjugate prior for brownain phylogenetic VCV likelihood

func InitNGPrior 

func InitNGPrior(mu0, k0, alpha0, beta0 float64) *NormalGammaPrior

InitNGPrior will initialize the prior on the matrix and BM mean

type SiteConfiguration 

type SiteConfiguration struct {
	Sites         map[int]map[int]bool
	AIC           float64
	ClusterTrees  map[int]string
	ClusterSizes  map[int]int
	ClusterString string
}

func (*SiteConfiguration) CalcClusterSizes 

func (c *SiteConfiguration) CalcClusterSizes()

func (*SiteConfiguration) Equals 

func (c *SiteConfiguration) Equals(check *SiteConfiguration) (equal bool)

type UVNDPPGibbs 

type UVNDPPGibbs struct {
	Dist            *DistanceMatrix
	Clusters        map[int]*UVNLike
	SiteAssignments map[int]int
	Prior           *NormalGammaPrior
	Gen             int
	PrintFreq       int
	WriteFreq       int
	Threads         int
	Workers         int
	Alpha           float64
	ClustOutFile    string
	LogOutFile      string
}

UVNDPPGibbs is a struct to store the attributes of a collapsed DPP gibbs sampler for the BM phylo gene expression data

func InitUVNGibbs 

func InitUVNGibbs(nodes []*Node, prior *NormalGammaPrior, gen, print, write, threads, workers int, alpha float64, treeOut, logOut string) *UVNDPPGibbs

InitUVNGibbs will initialize the parameters for the collapsed gibbs sampler.

func (*UVNDPPGibbs) ClusterString 

func (chain *UVNDPPGibbs) ClusterString() string

ClusterString will return a string of the current set of clusters

func (*UVNDPPGibbs) Run 

func (chain *UVNDPPGibbs) Run()

Run will run MCMC simulations

type UVNLike 

type UVNLike struct {
	Sites []int
	//MeanData *mat.Dense
	DataSum []float64
	KN      []float64
	MuN     []float64
	AlphaN  []float64
	BetaN   []float64
}

UVNLike will store all data and params for each cluster

Source Files

View all Source files

Directories

Path Synopsis

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.