Faster R: Why rowSums in R is so fast?

Question

I was curious about computation time in R especially while dealing with matrix or data frame. Let's notice the following example:

    slowFun = function() {
      mat = as.data.frame(matrix(rnorm(10000 * 10), 10000, 10))
      mat[, (ncol(mat) + 1)] = NA
      for (i in 1:nrow(mat)) {
        mat[i,ncol(mat)] = sum(mat[i,], na.rm = T)
      }
    }

    fastFun = function() {
      mat = as.data.frame(matrix(rnorm(10000 * 10), 10000, 10))
      mat[, (ncol(mat) + 1)] = NA
      mat[,ncol(mat)] = rowSums(mat, na.rm = T)
    }


    system.time(slowFun())
    system.time(fastFun())

   > system.time(slowFun())
   user  system elapsed 
   3.614   0.219   3.837 
   > system.time(fastFun())
   user  system elapsed 
   0.012   0.001   0.013

rowSums was impressive. When I tried to learn all about it:

    rowSums

It does not let me see how it was coded! I see .Internal and don't know where to look about it.
Any help will be appreciated.

This question is in the arena of the wider topic of vectorization in R. Of which you will find many discussions around. Here's one explanation: http://www.noamross.net/blog/2014/4/16/vectorization-in-r--why.html, especially part four on functions. — Pierre L, Sep 29 '15 at 02:53
http://stackoverflow.com/questions/14035506/how-to-see-the-source-code-of-r-internal-or-primitive-function/14035586#14035586 — Ben Bolker, Sep 29 '15 at 02:58
How to view the code for an `.Internal` function is described in the answer to [How can I view the source code for a function?](http://stackoverflow.com/q/19226816/271616). — Joshua Ulrich, Sep 29 '15 at 02:59
The reason is that the function was written in c++. If you are astounded by the speed of rowSums you should definitely try using Rfast's implementation (i.e. Rfast::rowsums) — Stefanos, Oct 10 '18 at 23:31

Faster R: Why rowSums in R is so fast?

0 Answers0