log_likelihood_Y: Marginal log-likelihood of Y

In Cronjaeger/HotBirdHMM: Find Recombination Hoitspots using a Hidden Markov Model

Description

Marginalizes out the hidden variables X using the forward algorithm to compute the log-likelihood of an observed sequence of emissions.

Usage

log_likelihood_Y(Y, M, rho = NULL, theta = NULL, Px = NULL, mu = NULL,
  prior_x0 = NULL)

Arguments

Y	observed sequence of emissions (1: heterozygous site; 0: homozygous site).
M	size of hidden statespace; an integer.
rho	the recombinartion rate between two ajacent sites.
theta	mutation rate between two sequences at any given position.
Px	transition kernel of hidden variables. May be passed instead of rho.
mu	a vector of emission probabilities. The ith entry is interpreted as the probability of Y=1 given X=i. May be passed instead of theta
x0_prior	initial distribution of the hidden markov chain. Defaults to a uniform distribution.

Value

a single number, corresponding to l(theta,rho|Y)

Cronjaeger/HotBirdHMM documentation built on May 9, 2019, 6:01 p.m.