wjakethompson/portableParallelSeeds: Allow Replication of Simulations on Parallel and Serial Computers

Creates seeds for simulation projects. Uses the L'Eucyer "many separate streams" approach to allow users to create one or more unique streams per simulation run. Can save, reload, and select among streams of random numbers for arbitrarily many projects. From a batch of thousands of simulations, one can select any arbitrary "run" of the model and re-initialize it (replicate it) in isolation. The book-keeping work is done in an enviroment called ".pps". This simulation will alter the state of the user session's .RandomSeed variable so as to re-initialize a simulation exactly. This was developed on a Linux cluster for replication of simulations across many compute nodes, but it is also compatible with iteration on a single computer. Includes vignettes to explain some basics of the package and the usage of pseudo-random number generators (PRNG).

Getting started

Package details

Maintainer
LicenseGPL (>= 2)
Version0.97
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
install.packages("remotes")
remotes::install_github("wjakethompson/portableParallelSeeds")
wjakethompson/portableParallelSeeds documentation built on Nov. 5, 2019, 12:22 p.m.