User:Timothee Flutre/Notebook/Postdoc/2011/11/09

Entry title

• in R, replace the row names of a matrix by their new names when the order is different (but assuming one-to-one mapping):

links <- data.frame(id1=c("a","b","c"), id2=c("1","2","3"), stringsAsFactors=FALSE)
mat <- matrix(runif(3*10), nrow=3)
rownames(mat) <- c("b","c","a")
new.rownames <- unlist(lapply(rownames(mat), function(i){links$id2[which(links$id1 == i)]}))
rownames(mat) <- new.rownames

Or it's maybe easier with the built-in match function:

new.rownames <- links$id2[ match(rownames(mat), links$id1) ]

• in R, visualize the missing values (encoded as "NA") in a matrix:

image(1:nrow(mat), 1:ncol(mat), is.na(mat), col=c("white","black"),
      main="Missing values", xlab="Genes", ylab="Samples")