added first Independent Q Learning experiment

[Mytolo]

PR Checklist

• Update NEWS.md?
• Unit tests for all structs / functions?
• Integration and correctness tests using a simple env?
• PR Review?
• Add or update documentation?
• Write docstrings for new methods?
  Here is an experiment for RLExperiments showing how to set up a simple independent learning with two players. (In the example, the roles of randomly choosing an action and taking create_policy -> DQN can be changed and they can also both try to approximate their Q values for action selection)

[codecov]

Codecov Report

Merging #922 (c744802) into main (8f5ea30) will increase coverage by 0.01%.
The diff coverage is 16.66%.

❗ Current head c744802 differs from pull request most recent head e5d3e6e. Consider uploading reports for the commit e5d3e6e to get more accurate results

@@           Coverage Diff            @@
##            main    #922      +/-   ##
========================================
+ Coverage   0.02%   0.04%   +0.01%     
========================================
  Files        215     209       -6     
  Lines       7622    7421     -201     
========================================
+ Hits           2       3       +1     
+ Misses      7620    7418     -202     

Files	Coverage Δ
...xperiments/src/ReinforcementLearningExperiments.jl	100.00% <100.00%> (ø)
...deps/experiments/experiments/MARL/pettingzoo_ex.jl	0.00% <0.00%> (ø)
...eps/experiments/experiments/MARL/IDQN_TicTacToe.jl	0.00% <0.00%> (ø)

... and 36 files with indirect coverage changes

[Mytolo]

@HenriDeh, @jeremiahpslewis would you have a look? It is supposed to just be an example on how to use multiAgentPolicy.

[HenriDeh]

I don't see a problem with the implementation but it needs to be added to the tests of RLExperiments like the other ones are. Otherwise we may not see that a change breaks this example and it will live there broken until someone notices.

[Mytolo]

I don't see a problem with the implementation but it needs to be added to the tests of RLExperiments like the other ones are. Otherwise we may not see that a change breaks this example and it will live there broken until someone notices.

I this what you intended to say here @HenriDeh ? Commit 82f77f4

[Mytolo]

FYI: rebased commit structure as it was messed up.

[HenriDeh]

Yes thank you. Gotta figure out why the CI fails now.

[Mytolo]

Finally, the pipeline should be correct. I needed to add some parts, but this is also helpful for other 3rd party environments once they need to be tested.

[Mytolo]

I don't see a problem with the implementation but it needs to be added to the tests of RLExperiments like the other ones are. Otherwise we may not see that a change breaks this example and it will live there broken until someone notices.

this should be solved now @HenriDeh ?

[HenriDeh]

Hi @Mytolo, sorry for the delay, I was on vacations. You have conflicts to resolve then I can approve the PR for merging :)