Collaborating on the project

[hdmetor]

Hi, I'd like to collaborate on this project so that HLearn gets usable.

Do you have a list of issues that need contributions? I'm looking to start with something simple.

[mikeizbicki]

There are two options:

1. Find some FIXME in the source code that sounds fun to you and fix that. You can see Add axioms of intuitionistic logic to Heyting algebras #41 as an example.
2. Add support for sparse linear algebra. This will involve writing bindings to a BLAS library for sparse linear algebra. There are currently three implementations of linear algebra in subhask. The first thing to look at is is this example which gives an overview of the linear algebra operations. This example uses the HMatrix bindings. HMatrix is a Haskell library that provides bindings to standard BLAS libraries. Two other implementations are the accelerate GPU backend currently under development (see Initial implementations of Accelerate backends for Vector and Matrix #39), and a native haskell backend that is slow but does not require dependencies (see Algebra.Matrix #32).

[hdmetor]

Thanks for the answer. Will have a look and come back with some extra questions!

[tonyday567]

There's a lot of undefined's also - filling these in or adding a fixme if it's unclear would be an easy entry to helping out.