The goal of wrswoR is to provide faster implementations of weighted random sampling without replacement in R.
You can install the released version of wrswoR from CRAN with:
install.packages("wrswoR")
And the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("krlmlr/wrswoR")
The functions in this package are a drop-in replacement to
sample.int(n, size, replace = FALSE, prob = prob)
. With large n
,
sample.int()
becomes too slow to be practical, unlike the functions in
this package.
library(wrswoR)
set.seed(20200726)
sample_int_crank(20, 10, 1:20)
#> [1] 8 18 14 17 11 15 10 4 13 5