Added Python efr algorithm implementation

[Jamesflynn1]

A Python implementation of the EFR (https://arxiv.org/abs/2102.06973) algorithm with the deviation types defined in the proposing paper. The implementation was developed as part of my undergraduate dissertation and I thought it would be beneficial if the implementation could be made available as part of OpenSpiel.

Uses a shortcut for external only deviations as found in https://github.com/dmorrill10/hr_edl_experiments (original experimental code)

Any guidance or feedback would be appreciated as this is my first contribution to an open source project.

Todo:

• Look into issues with behavioural deviations
• Look into possible incorrect definition with CSPS deviations
• Add docstrings for all functions
• Remove discounting code
• Refactor code
• Ensure adherence to Google Python style
• Add updates to OpenSpiel docs

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[lanctot]

Thanks @Jamesflynn1 !

Please let us know when it is ready for a review.

[lanctot]

Please fix the style to conform to Google Python style. I've pointed out a few examples, but please fix them all.

Use pylint with the pylintrc from the Google style guide

[lanctot]

Hi @Jamesflynn1, if any of your recent commits have addressed the comments, can you reply to them and/or mark them as resolved?

[Jamesflynn1]

Hi @Jamesflynn1, if any of your recent commits have addressed the comments, can you reply to them and/or mark them as resolved?

Hi @lanctot, apologies I haven't replied sooner, I have been very busy with university coursework and graduate applications.

I ran pylint with the Google style guide and I was still some of style issues (spaces after comments and blank lines), I'll take another look now and mark off the solved issues.

[Jamesflynn1]

Hi @lanctot,

I have tried to refractor the files using the linked pylintrc file and I'm not getting all the commenting suggestions that has been pointed out. I've fixed (hopefully) all instances of the issues you have pointed out. However, I think it's possible that I have a different pylintrc file (I get no linting issues when I pylint the file) so to catch all issues would it be possible to compare files?

Any assistance would be appreciated!

Thanks,
James

[lanctot]

No problem, I can try to fix these on our side. Apologies for the delays.

[lanctot]

Hi @Jamesflynn1 , sorry for the extreme lateness. Did you ever verify the convergence matches what was expected from the papers? I'd like to import this, but it would be good to know that it's correct.

[lanctot]

Can you add efr_test.py to the python/CMakeLists.txt

We need to ensure that it runs on Github Actions CI before we import it.

You might need to pull changes from master because that file has changed a few times since our last contact.