Fast Pseudo Random Number Generators

Several fast random number generators are provided as C++ header only libraries: The PCG family by O'Neill (2014 < https://www.cs.hmc.edu/tr/hmc-cs-2014-0905.pdf>) as well as Xoroshiro128+ and Xoshiro256+ by Blackman and Vigna (2018 ). In addition fast functions for generating random numbers according to a uniform, normal and exponential distribution are included. The latter two use the Ziggurat algorithm originally proposed by Marsaglia and Tsang (2000, ). These functions are exported to R and as a C++ interface and are enabled for use with the default 64 bit generator from the PCG family, Xoroshiro128+ and Xoshiro256+ as well as the 64 bit version of the 20 rounds Threefry engine (Salmon et al., 2011 ) as provided by the package 'sitmo'.

The dqrng package provides fast random number generators (RNG) with good statistical properties for usage with R. It combines these RNGs with fast distribution functions to sample from uniform, normal or exponential distributions. Both the RNGs and the distribution functions are distributed as C++ header-only library.

Installation

The currently released version is available from CRAN via

install.packages("dqrng")

Intermediate releases can also be obtained via drat:

if (!requireNamespace("drat", quietly = TRUE)) install.packages("drat")
drat::addRepo("daqana")
install.packages("dqrng")

Example

Using the provided RNGs from R is deliberately similar to using R’s build-in RNGs:

library(dqrng)
dqset.seed(42)
dqrunif(5, min = 2, max = 10)
#> [1] 9.211802 2.616041 6.236331 4.588535 5.764814
dqrexp(5, rate = 4)
#> [1] 0.35118613 0.17656197 0.06844976 0.16984095 0.10096744

They are quite a bit faster, though:

N <- 1e4
bm <- bench::mark(rnorm(N), dqrnorm(N), check = FALSE)
bm[, 1:5]
#> # A tibble: 2 x 5
#>   expression      min     mean   median      max
#>   <chr>      <bch:tm> <bch:tm> <bch:tm> <bch:tm>
#> 1 rnorm(N)      657µs  752.4µs  727.5µs   1.09ms
#> 2 dqrnorm(N)     72µs   85.8µs   80.8µs 166.02µs

This is also true for the provided sampling functions with replacement:

m <- 1e7
n <- 1e5
bm <- bench::mark(sample.int(m, n, replace = TRUE),
                  sample.int(1e3*m, n, replace = TRUE),
                  dqsample.int(m, n, replace = TRUE),
                  dqsample.int(1e3*m, n, replace = TRUE),
                  check = FALSE)
bm[, 1:5]
#> # A tibble: 4 x 5
#>   expression                                 min     mean   median      max
#>   <chr>                                 <bch:tm> <bch:tm> <bch:tm> <bch:tm>
#> 1 sample.int(m, n, replace = TRUE)      905.05µs   1.11ms   1.08ms   1.81ms
#> 2 sample.int(1000 * m, n, replace = TR…   1.69ms   1.97ms   1.92ms   2.85ms
#> 3 dqsample.int(m, n, replace = TRUE)    274.76µs 333.97µs 315.47µs 604.48µs
#> 4 dqsample.int(1000 * m, n, replace = … 340.61µs 413.71µs 377.36µs 888.39µs

And without replacement:

bm <- bench::mark(sample.int(m, n),
                  sample.int(1e3*m, n),
                  sample.int(m, n, useHash = TRUE),
                  dqsample.int(m, n),
                  dqsample.int(1e3*m, n),
                  check = FALSE)
bm[, 1:5]
#> # A tibble: 5 x 5
#>   expression                            min     mean   median      max
#>   <chr>                            <bch:tm> <bch:tm> <bch:tm> <bch:tm>
#> 1 sample.int(m, n)                  21.97ms  21.97ms  21.97ms  21.97ms
#> 2 sample.int(1000 * m, n)            5.21ms   6.34ms   5.78ms  11.28ms
#> 3 sample.int(m, n, useHash = TRUE)   3.25ms   3.97ms   3.61ms   8.43ms
#> 4 dqsample.int(m, n)                  1.2ms   1.62ms    1.4ms   4.37ms
#> 5 dqsample.int(1000 * m, n)          1.77ms   2.32ms   2.09ms   4.87ms

Note that sampling from 10^10 elements triggers “long-vector support” in R.

In addition the RNGs provide support for multiple independent streams for parallel usage:

N <- 1e7
dqset.seed(42, 1)
u1 <- dqrunif(N)
dqset.seed(42, 2)
u2 <- dqrunif(N)
cor(u1, u2)
#> [1] -0.0005787967

Feedback

All feedback (bug reports, security issues, feature requests, …) should be provided as issues.