Weighted Random Sampling without Replacement
A collection of implementations of classical and novel
algorithms for weighted sampling without replacement.
A package with different implementations of weighted random sampling without replacement in R.
wrswoR 1.1 (2018-02-02)
- Use microbenchmark package conditionally.
- Remove dependency on cluster.
- Use native method registration.
- Prepare for submission to R Journal.
- Improve vignette.
- Internal tweaks.
Version 1.0-1 (2016-02-26)
- Finally submitted version of the paper.
Version 1.0 (2016-02-22)
First CRAN release.
- Alternative implementations for weighted random sampling, as implemented by R's
sample.int(replace = FALSE, prob = ...)
sample_int_rej(): A rejective algorithm, in R
sample_int_rank(): One-pass sampling by Efraimidis and Spirakis, in R
sample_int_crank(): One-pass sampling by Efraimidis and Spirakis, in C++
sample_int_expj(): Reservoir sampling with exponential jumps Efraimidis and Spirakis, in C++
- Two more experimental functions.
- Article for submission to JStatSoft as vignette.