initBdHMM: Initialization of bidirectional hidden Markov models

In STAN: The Genomic STate ANnotation Package

Description

Initialization of bidirectional hidden Markov models

Usage

initBdHMM(obs, dStates = 0, uStates = 0, method, dirFlags = NULL, directedObs = rep(0, ncol(obs[[1]])), sizeFactors = matrix(1, nrow = length(obs), ncol = ncol(obs[[1]])), sharedCov = FALSE)

Arguments

obs	The observations. A list of one or more entries containing the observation matrix (numeric) for the samples (e.g. chromosomes).
dStates	The number of directed states.
uStates	The number of undirected states.
method	Emission distribution of the model. One out of c("NegativeBinomial", "PoissonLogNormal", "NegativeMultinomial", "ZINegativeBinomial", "Poisson", "Bernoulli", "Gaussian", "IndependentGaussian")
dirFlags	The flag sequence is needed when a bdHMM is fitted on undirected data (e.g.) ChIP only. It is a list of character vectors indication for each position its knwon directionality. U allows all states. F allows undirected states and states in forward direction. R allows undirected states and states in reverse direction.
directedObs	Integer vector defining the directionality (or strand-specificity) of the data tracks. Undirected (non-strand-specific) data tracks (e.g. ChIP) are indicated by '0'. Directed (strand-specific) data tracks are indicated by increasing pairs of integers. For instance c(0,0,0,1,1,2,2): The first three data tracks are undirected, followed by two pairs of directed measurements.
sizeFactors	Library size factors for Emissions PoissonLogNormal or NegativeBinomial as a length(obs) x ncol(obs[[1]]) matrix.
sharedCov	If TRUE, (co-)variance of (Independent)Gaussian is shared over states. Only applicable to 'Gaussian' or 'IndependentGaussian' emissions. Default: FALSE.

Value

A HMM object.

Examples

data(example)
bdHMM_ex = initBdHMM(observations, dStates=3, method="Gaussian") 

STAN documentation built on Nov. 8, 2020, 11:11 p.m.