Number of Kills for each unit and team feature request

[kmcnayr]

I am interested in seeing the number of kills each unit and team achieve in the battle. It seems like keeping the running total in M might work. The unit level totals would be available in the last frame of the simulation which you can get by running with return_frames = True.

In the return_frames = False scenario I suppose a more robust results dataset would need to be implemented which in addition to returning the remaining units for each team it would also include the number of kills each team achieved. If this were to be done a pandas dataframe with the teams on the rows and statistics in the columns seems ideal. Other stats could also be added like total damage done, team level accuracy, units lost, types of units killed, etc.

Obviously this is not a trivial enhancement but wanted to note it here.

[gregparkes]

Thank you for the issue.

These are all great suggestions. If you have time I am open to review code from PRs. However, I don't currently have time to commit development to this repo.

As it stands, there are a number of glaring issues, particularly as this was written 5+ years ago in numba, I think now I would port most of the code into Cython and include a compile step in the setup.py.

Another alternative would be to create a Gym environment to enable reinforcement learning users to have a go.