# loglikelihood: The (incomplete data) loglikelihood function for the negative... In johannabertl/SelectionMix: Negative binomial mixture model

## Description

The (incomplete data) loglikelihood function for the negative binomial mixture model

## Usage

 `1` ```loglikelihood(theta, x.syn, x.non, cvec) ```

## Arguments

 `theta` c(alpha, beta, p1, p2, ..., pk-1) `x.syn` vector of synonymous mutation count `x.non` vector of non-synonymous mutation count `cvec` vector of k positive values c(c1, ..., ck)

Johanna Bertl

## Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31``` ```c1 = 0.5; c2 = 10 p1 = 0.2; p2 = 0.8 alpha = 10 beta = 5 mutations = rnegbinmix_syn_nonsyn(n=3000, alpha=alpha, beta=beta, c=c(c1, c2), p = c(p1, p2)) # log likelihood at the true value loglikelihood(theta=c(alpha,beta,p1), x.syn = mutations\$Syn, x.non = mutations\$Non, cvec=c(c1, c2)) # likelihood surface alphavec = 1:20 betavec = seq(0.5, 10, by=0.5) p1vec = seq(0.11, 0.3, by=0.01) llalpha = numeric(20) llbeta = numeric(20) llp1 = numeric(20) for(i in 1:20){ llalpha[i] = loglikelihood(theta=c(alphavec[i],beta,p1), x.syn = mutations\$Syn, x.non = mutations\$Non, cvec=c(c1, c2)) llbeta[i] = loglikelihood(theta=c(alpha,betavec[i],p1), x.syn = mutations\$Syn, x.non = mutations\$Non, cvec=c(c1, c2)) llp1[i] = loglikelihood(theta=c(alpha,beta,p1vec[i]), x.syn = mutations\$Syn, x.non = mutations\$Non, cvec=c(c1, c2)) } plot(alphavec, llalpha, t="b") abline(v=alpha, col="red") plot(betavec, llbeta, t="b") abline(v=beta, col="red") plot(p1vec, llp1, t="b") abline(v=p1, col="red") ```

johannabertl/SelectionMix documentation built on May 3, 2019, 4:03 p.m.