A new release for GNU Octave

Re:A new release for GNU Octave by Frost1x on Sunday January 03, 2021 @09:14AM (#25621862)

This is the main use case where I've used octave: working with someone of a group who has developed fairly involved Matlab codebase and we need to make it more scalable in some way, either for long running computation, reducing license costs, perform distributed memory work, or enable more robust parametric space exploration.

The Achilles heel here is that in Matlab, there's quite often a licensed toolbox dependency which makes portability more challenging. Every now and again I've seen pure Matlab code and the Octave transition is smooth. Sometimes there are so many toolbox dependencies that it just isn't even worth trying.

[ Parent | on HN ]

Re:A new release for GNU Octave by bayindirh on Sunday January 03, 2021 @09:35AM (#25621979)

The luck in my case is, since we're testing either algorithms for formulations, most (if not all) code is written in pure MATLAB and has no toolbox dependency.

However, we're living in a matrix world. MATLAB is slow and Octave is even slower while working with matrices. I have no idea how to fix that part.

[ Parent | on HN ]

Re:A new release for GNU Octave by Jtsummers on Sunday January 03, 2021 @11:40AM (#25622808)

> MATLAB is slow and Octave is even slower while working with matrices.

I'm curious what you mean with this. Matlab was always fast with matrices (but my experience is now from 2 decades ago). The thing that was slow was loops. That is:

  for i = 1:100
    c(i) = a(i) + b(i);
  end

would always be slower than this:

  c = a .+ b;

The trick in my work was to always find matrix/vector operations that were equivalent to my desired loop and conditional structures. I could improve execution by 100x or more for many programs doing this.

[ Parent | on HN ]