Search code examples
rmulticore

speed up gls in R?

I have a Intel i7 quad-core processor. Is it possible to use multiple cores at the same time to calculate a single gls model? These can take as much as 5-10 minutes to run, but they only use one core at a time.

If I have four models that take a few minutes each, would it be better to run each model on a separate core?

What options do I have?

Solution

  • You have a couple options:

    1. Run your models simultaneously by using a package like parallel, foreach, snow, multicore, etc.
    2. Get a build of R with a multi-threaded BLAS.

    Note that you should do one or the other of these, but not both at the same time.