Fastest way to sort each row of a large matrix in R

Question

I have a large matrix:

set.seed(1)
a <- matrix(runif(9e+07),ncol=300)

I want to sort each row in the matrix:

> system         


        

Answer 1

Another excellent method from Martin Morgan without any usage of external packages in Fastest way to select i-th highest value from row and assign to new column:

matrix(a[order(row(a), a)], ncol=ncol(a))

There is also an equivalent for sorting by columns under comments in the same link.

Timing code using same data as Craig:

set.seed(1)
a <- matrix(runif(9e7),ncol=300)

use_for <- function(){
    sorted3 <- a
    for(i in seq_len(nrow(a))) 
        sorted3[i,] <- sort.int(a[i,], method='quick') 
    sorted3
}

microbenchmark::microbenchmark(times=3L,
    t(apply(a,1,sort)),
    t(apply(a,1,sort.int, method='quick')),
    use_for(),
    Rfast::rowSort(a),
    t(apply(a,1,grr::sort2)),
    matrix(a[order(row(a), a)], ncol=ncol(a))
)

Timings:

Unit: seconds
                                        expr       min        lq      mean    median        uq       max neval
                        t(apply(a, 1, sort)) 37.875665 40.143190 41.098627 42.410715 42.710108 43.009502     3
  t(apply(a, 1, sort.int, method = "quick")) 26.406063 27.146861 27.714226 27.887659 28.368307 28.848955     3
                                   use_for() 20.038295 20.140692 20.504223 20.243088 20.737187 21.231285     3
                           Rfast::rowSort(a)  6.105679  6.460003  6.836455  6.814326  7.201844  7.589361     3
                  t(apply(a, 1, grr::sort2)) 11.912422 13.035231 13.667377 14.158040 14.544854 14.931669     3
 matrix(a[order(row(a), a)], ncol = ncol(a)) 10.307094 10.789946 11.294119 11.272797 11.787632 12.302466     3

---

And to present a more complete picture, another test for character class (excluding Rfast::rowSort as it cannot handle character class):

set.seed(1)
a <- matrix(sample(letters, 9e6, TRUE),ncol=300)

microbenchmark::microbenchmark(times=1L,
    t(apply(a,1,sort)),
    t(apply(a,1,sort.int, method='quick')),
    use_for(),
    #Rfast::rowSort(a),
    t(apply(a,1,grr::sort2)),
    matrix(a[order(row(a), a, method="radix")], ncol=ncol(a))
)

Timings:

Unit: milliseconds
                                                          expr        min         lq       mean     median         uq        max neval
                                          t(apply(a, 1, sort)) 30392.7951 30392.7951 30392.7951 30392.7951 30392.7951 30392.7951     1
                    t(apply(a, 1, sort.int, method = "quick")) 29359.7711 29359.7711 29359.7711 29359.7711 29359.7711 29359.7711     1
                                                     use_for() 31018.8827 31018.8827 31018.8827 31018.8827 31018.8827 31018.8827     1
                                    t(apply(a, 1, grr::sort2))  2539.1711  2539.1711  2539.1711  2539.1711  2539.1711  2539.1711     1
 matrix(a[order(row(a), a, method = "radix")], ncol = ncol(a))   480.7405   480.7405   480.7405   480.7405   480.7405   480.7405     1

---

Head to head:

set.seed(1)
a <- matrix(sample(letters, 9e7, TRUE),ncol=300)
microbenchmark::microbenchmark(times=1L,
    t(apply(a,1,grr::sort2)),
    matrix(a[order(row(a), a, method="radix")], ncol=ncol(a))
)

Timings:

Unit: seconds
                                                          expr       min        lq      mean    median        uq       max neval
                                    t(apply(a, 1, grr::sort2)) 29.098726 29.098726 29.098726 29.098726 29.098726 29.098726     1
 matrix(a[order(row(a), a, method = "radix")], ncol = ncol(a))  7.067744  7.067744  7.067744  7.067744  7.067744  7.067744     1