Re: [R] Expanding matrix into dummies

Thank you very much, Marc and Bert - this is great! On Tue, Dec 22, 2015 at 7:14 PM, Bert Gunter wrote: > ... Perhaps worth noting is that the row indices can be created > directly without row(): > > result[cbind(rep.int(seq_len(6),5), as.vector(x))] <- 1 > > but the downside is that you have to

Re: [R] Expanding matrix into dummies

... Perhaps worth noting is that the row indices can be created directly without row(): result[cbind(rep.int(seq_len(6),5), as.vector(x))] <- 1 but the downside is that you have to know that a matrix is a vector "stored" in column major order. I find this arcane detail quite handy, though. Cheer

Re: [R] Expanding matrix into dummies

> On Dec 22, 2015, at 4:42 PM, Dimitri Liakhovitski > wrote: > > # I have a matrix x: > > k <- 20 > N <- 5 > set.seed(123) > x <- matrix(c(sample(1:k, N, replace = F), > sample(1:k, N, replace = F), > sample(1:k, N, replace = F), > sample(1:k, N, replace

Re: [R] Expanding matrix

Hi,You could do: m1 <- as.matrix(read.table(text="3 0.9 1 5 0.5 0.7 7 0.1 0.3",sep="",header=FALSE)) dimnames(m1) <- NULL d2 <- data.frame(V1=1:10) d1 <- as.data.frame(m1) library(zoo) res <- na.locf(merge(d1,d2,all=TRUE)) res[is.na(res)] <- 1  res #or vec1 <- 1:10 indx <- vec1[!vec1 %in% m1[,1