Read predictive quantities output from BGmix.

Description

Reads predictive p-values from files output from BGmix. Also (optionally) reads posterior predictive distributions of data.

Usage

1
ccPred(filedir, q.partial = T, q.trace = F, quiet = T)

Arguments

filedir

character. The name of the output directory created by BGmix.

q.partial

logical. Read partial predictive p-values?

q.trace

logical. Read posterior predictive distributions of data?

quiet

logical. Parameter passed to 'scan'. (If false, 'scan' prints details of number of items read in.)

Value

pval.ss.post

matrices no. genes x no. conditions. Posterior predictive p-values for sum of squares for each gene in each condition.

pval.ss.mix

matrices no. genes x no. conditions. Mixed predictive p-values for sum of squares for each gene in each condition.

pval.ss.part

matrices no. genes x no. conditions. Partial predictive p-values for sum of squares for each gene in each condition.

pval.ybar.post

matrices no. genes x no. mixture components. Posterior predictive p-values for ybar for each gene in each mixture component.

pval.ybar.mix2

matrices no. genes x no. mixture components. Mixed predictive p-values for ybar for each gene in each mixture component.

pval.ybar.part

matrices no. genes x no. mixture components. Partial predictive p-values for ybar for each gene in each mixture component.

ybar.pred1

Posterior predictive distribution of ybar.

ybar.pred3

Mixed predictive distribution of ybar.

ss.pred1

Posterior predictive distribution of sums of squares.

ss.pred2

Mixed predictive distribution of sums of squares.

Note

Additional output: pval.ybar.mix1 and pval.ybar.mix3 are alternative versions of mixed predictive p-values (currently not used). Also, ybar.pred2 and ybar.pred4 are the corresponding alternative mixed predictive distributions for ybar.

Author(s)

Alex Lewin

Examples

1
2
3
4
5
## Note this is a very short MCMC run!
## For good analysis need proper burn-in period.
data(ybar,ss)
outdir <- BGmix(ybar, ss, c(8,8), nburn=0, niter=100, nthin=1)
pred <- ccPred(outdir)