R: Difference between revisions

For more on submitting jobs, see the [[Running jobs]] page.

Foreach can be considered as a unified interface for all backends (i.e. doMC, doMPI, doParallel, doRedis, etc.). It works on all platforms, assuming that the backend works. doParallel acts as an interface between foreach and the parallel package and can be loaded alone. There are some known efficiency issues when using foreach to run a very large number of very small tasks. Therefore, keep in mind that the following code is not the best example of an optimized use of the foreach() call but rather that the function chosen was kept at a minimum for demonstration purposes.

# to register the backend;

# to call foreach() by keeping it on the same line as the %do% (serial) or %dopar% operator.

1. Place your R code in a script file, in this case the file is called ''test_foreach.R''.

<!--T:51-->

For more on submitting jobs, see the [[Running jobs]] page.