Some examples of where explicitly using an integer scalar/vector is more efficient in R. Often the difference in a single operation is quite small, but if repeatedly using integer values in a script using explicit typing can be an easy way to gain speed.

integer_typing.R

library(microbenchmark)

# Examples where explictly typing with "L" is more efficient

## create a reasonably large matrix
r <- 5000
c <- 100
ints <- sample(0:10, r*c, replace = TRUE)
M <- matrix(ints, nr = r, nc = c)

## comparison to integer scalar
microbenchmark(M > 0, M > 0L, unit = "ms")

## comparison to integer vector
microbenchmark(M >= rep.int(1, r), M >= rep.int(1L, r), unit = "ms")

# Example where integer vector is more efficient

## index might originally come from a string, e.g. group label
grp_char <- sample(as.character(1:10), 10000, replace = TRUE)
num <- rnorm(10000)

## rowsum fastest with integer
## (conversion takes time, but if only need to do once in script where
## index used a lot, can gain overall)
grp_int <- as.integer(grp_char)
grp_num <- as.numeric(grp_char)
grp_fac <- as.factor(grp_char)
microbenchmark(rowsum(num, grp_char),
               rowsum(num, grp_num),
               rowsum(num, grp_fac),
               rowsum(num, grp_int), unit = "ms")

Interesting!