approx.expected.info {edgeR} | R Documentation |
Using a linear fit (for simplicity), the expected information from the conditional log likelihood of the dispersion parameter of the negative binomial is calculated over all genes.
approx.expected.info(object, d, qA, robust = FALSE)
object |
DGEList object containing the raw data with elements data (table of counts), group (vector indicating group) and lib.size (vector of library sizes) |
d |
delta parameter for negative binomial - phi/(phi+1) |
qA |
list from output of quantileAdjust |
robust |
logical on whether to use a robust fit, default FALSE |
vector of Fisher information approximates (with length same as the number of rows of the original data)
Mark Robinson
set.seed(0) y<-matrix(rnbinom(40,size=1,mu=10),ncol=4) d<-list(data=y,group=rep(1:2,each=2),lib.size=rep(c(1000:1001),2)) qA<-quantileAdjust(d,alpha=100) exp.inf<-approx.expected.info(d,1/(1 + qA$r[1]),qA)