search

SebastianRiedel / bingham

Bingham Statistics Library (BSL)
BSD 3-Clause "New" or "Revised" License
42 stars 21 forks source link

Python bindings #11

Open adamconkey opened 2 years ago

adamconkey commented 2 years ago

This library seems to be the most comprehensive Bingham library out there, but I was sad to find it was only in C and Matlab. I have started making some Python bindings on my fork (building off of what was done on this fork). I'm currently creating a minimal interface for only what I need in my research currently, but am considering circling back in the coming months to start working on a more comprehensive set of bindings. Would this be a welcome contribution, and would there be someone to review a PR when complete? It's unclear to me how actively maintained this library is/will be. Thanks!

jglov3id commented 2 years ago

Hi Adam,

That would be great, thanks! I can try to do the PR although there may be delays if work is busy.

Best, Jared

On Tue, Mar 22, 2022 at 1:21 PM Adam Conkey @.***> wrote:

This library seems to be the most comprehensive Bingham library out there, but I was sad to find it was only in C and Matlab. I have started making some Python bindings on my fork https://github.com/adamconkey/bingham (building off of what was done on this fork https://github.com/r-pad/bingham). I'm currently creating a minimal interface for only what I need in my research currently, but am considering circling back in the coming months to start working on a more comprehensive set of bindings. Would this be a welcome contribution, and would there be someone to review a PR when complete? It's unclear to me how actively maintained this library is/will be. Thanks!

— Reply to this email directly, view it on GitHub https://github.com/SebastianRiedel/bingham/issues/11, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABY3HAP72PGALPPFLVNX2EDVBH6QBANCNFSM5RLQIRGQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

adigitoleo commented 1 year ago

I would also be interested in a Python interface. In particular I'd like to try parametric density estimation. I believe that nuklei implements non-parametric density estimation only, just based on their documentation.

@adamconkey I wonder if the code on your fork can already be used to do this? Could you perhaps give an example usage if you have time? I'm not experienced with spherical statistics (and generally a bit rusty on stats). As input I would have a set of points on a unit sphere.