Mdn cleanup/deprecating hidden features

[anasbekheit]

Deprecated parameter hidden_features in mdn.py

Problem Description

Previously users were required to pass in hidden_features which is un-necessary and introduces the following problems

• Users are required to keep track of their hidden_net output in their code in order to pass it properly which adds unnecessary overhead on the user.
• In the case that the user passes a wrong hidden_features that doesn't match hidden_net output the code fails.

Proposed Solution

Instead of requiring the user to provide hidden_features, the code should infer hidden_features from hidden_net.

Other Considerations

Code has to maintain backwards compatibility with code that is maybe using hidden_features as a positional argument.

This is done by converting hidden_features, hidden_net, num_components into kwargs and manually checking for their existence.

Misc

added .idea/ to .gitignore to support Idea IDEs

[janfb]

Thanks for contributing this @anasbekheit !

It all looks good. I just made a couple of suggestions: keep hidden_net and num_components as required args; use hidden_features when it is passed and make sure it fits.

Once those are in, it's good to be merged.
Thanks again 🙏

[janfb]

Another comment: We are planning to re-license pyknos under the Apache 2.0 license, see #28 .

Please comment under this PR that you approve this re-licensing, otherwise we will not be able to merge the PR.

Thanks!

[anasbekheit]

Another comment: We are planning to re-license pyknos under the Apache 2.0 license, see #28 .

Please comment under this PR that you approve this re-licensing, otherwise we will not be able to merge the PR.

Thanks!

Sure, no worries! I approve of the re-licensing of the code.

[anasbekheit]

Thanks for contributing this @anasbekheit !

It all looks good. I just made a couple of suggestions: keep hidden_net and num_components as required args; use hidden_features when it is passed and make sure it fits.

Once those are in, it's good to be merged. Thanks again 🙏

Hi @janfb, thanks for taking the time to review, I committed your suggestions, let me know if anything else needs polishing.

Cheers!

[janfb]

Yes @anasbekheit , could you please merge main into your branch to resolve the merge conflicts?
That'd be great 🙏

[anasbekheit]

@janfb

Yes @anasbekheit , could you please merge main into your branch to resolve the merge conflicts? That'd be great 🙏

I am not sure I understand, I did a git pull, there were no conflicts with main.

[janfb]

@janfb

Yes @anasbekheit , could you please merge main into your branch to resolve the merge conflicts? That'd be great 🙏

I am not sure I understand, I did a git pull, there were no conflicts with main.

My bad, only the rebasing would not work due to your changes. Squashing works just fine.
Thanks!