NEWS.md

In wrswoR: Weighted Random Sampling without Replacement

wrswoR 1.1.1 (2020-07-26)

• In sample_int_expj(), move retrieval of top element into the loop, to make it independent from the actual implementation of the priority queue.

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.

wrswoR documentation built on July 27, 2020, 1:08 a.m.