Open luiscarlosgph opened 7 years ago
I don't think that you have continuous integration on travis (or circle). We are not requiring it for integration in scikit-learn-contrib (I think that we should, @mblondel, @amueller, any opinion), but I think that it would be much better for your project.
I'm +1 for adding a requirement for CI. (maybe at least a soft requirement.) It's not so difficult to set up as long as the test suite already exists, and we do require tests.
I had subtle bugs in polylearn because of using non-initialized memory in cython which only showed up on 32-bit windows thanks to Appveyor. Would never have caught them without CI & thorough tests.
Thanks for the quick feedback, I will add the CI with Travis now and come back to you when done.
It's the last point on the the workflow document:
Where did you get from that we don't require it? If that requirement (and I think it is a requirement) is not documented in a particular place, we need to add it.
CI with Travis: https://travis-ci.org/luiscarlosgph/t-Student-Mixture-Models
Where did you get from that we don't require it?
OK. I guess we are all on the page. I was maybe trying to be too careful/polite.
With regards to integrating this, it's a pity that the docs are not visible online. Is there a plan to change this after integration?
I have put the documentation available here: http://t-student-mixture-models.readthedocs.io/en/latest
Request for project inclusion in scikit-learn-contrib