13

I'm trying to port some Python code to Scala. It makes heavy use of Numpy and Scipy. While I've found a number of dense matrix / linear algebra libraries that will do as an adequate (but not superb) replacement for NumPy, I've not really found anything that provides the functionality I use in SciPy. In particular, I'm looking for a libraries that support sparse partial eigendecompositions (like SciPy's wrapping of arpack), and then libraries for some simple things that SciPy provides (e.g. histograms).

nomad
  • 1,809
  • 2
  • 18
  • 33
  • Even though I'm not trying to suggest that this question is on the wrong forum; you might also try reposting at http://cstheory.stackexchange.com/ just because there are a lot of people who do grad work and research there; its likely a portion of them have familiarity with mathematical programming libraries. – Dave Apr 10 '11 at 16:01
  • 2
    Just curious: why have you decided to migrate from Python to Scala? I'm asking because I'm so far a Python person as well (amongst many other languages) and now actively learning Scala. – Erik Kaplun Oct 01 '13 at 09:47

2 Answers2

17

There have been a few similar questions already on this topic:

  1. Java/Scala library for algebra, mathematics
  2. Java / Scala math library with elliptic integrals and bessel functions?
  3. Clojure or Scala for bioinformatics/biostatistics/medical research

Also, http://code.google.com/p/scalalab/ looks interesting.

Community
  • 1
  • 1
Dean Wampler
  • 2,141
  • 13
  • 10
2

Blatant opportunism: I'm the author of numsca, the numpy for scala. Obviously the best in town.

Numsca lives here: https://github.com/botkop/numsca

Feedback appreciated.

Thank you.

botkop
  • 934
  • 1
  • 8
  • 17