Jorge, I have not used it myself, but you might find the dist() function in the proxy package to be useful.
http://cran.r-project.org/web/packages/proxy/index.html Jean On Mon, Jan 19, 2015 at 7:38 AM, Jorge I Velez <jorgeivanve...@gmail.com> wrote: > Dear all, > > Given vectors "x" and "y", I would like to compute the proportion of > entries that are equal, that is, mean(x == y). > > Now, suppose I have the following matrix: > > n <- 1e2 > m <- 1e4 > X <- matrix(sample(0:2, m*n, replace = TRUE), ncol = m) > > I am interested in calculating the above proportion for every pairwise > combination of rows. I came up with the following: > > myd <- function(X, p = NROW(X)){ > D <- matrix(NA, p, p) > for(i in 1:p) for(j in 1:p) if(i > j) D[i, j] <- mean(X[i, ] == X[j,]) > D > } > > system.time(d <- myd(X)) > > However, in my application n and m are much more larger than in this > example and the computational time might be an issue. I would very much > appreciate any suggestions on how to speed the "myd" function. > > Note: I have done some experiments with the dist() function and despite > being much, much, much faster than "myd", none of the default distances > fits my needs. I would also appreciate any suggestions on how to include > "my own" distance function in dist(). > > Thank you very much for your time. > > Best regards, > Jorge Velez.- > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide > http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. > [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.