Add a long-run strategy to the filtering test

[caddycarine]

Currently test_boolean_filtering only uses strategies as given by the strategy_list hypothesis strategy. This by default only picks short run time strategies so we are not ensuring that the long run filter works.

[Nikoleta-v3]

Hello @caddycarine 👋 Thank you for the PR. Nice catch that the long run strategies are currently not included in the test.

The DBS strategy is indeed a long run strategy but Cooperator is not. Cooperator is actually included in the strategy_list.

Note that you can see the list of the long run strategies using the following code:

>>> import axelrod as axl

>>> filterset = {
...    "long_run_time": True
...    }

>>> strategies = axl.filtered_strategies(filterset)
>>> strategies

which is based on this page of the tutorial: https://axelrod.readthedocs.io/en/stable/tutorials/advanced/classification_of_strategies.html?highlight=classifier 😄 hope this helps.

[drvinceknight]

I think that's wanted behaviour here @Nikoleta-v3, this test is checking that if you pass a list of strategies, some of which are long run time then the filter will pick the long run time ones out. Whether or not cooperator is long run time is not terribly important, it just needed an example with at least 1 long run time strategy. Am I missing something?

[drvinceknight]

@caddycarine the tests are currently failing but this is not anything you've done (probably a dependency that's been upgraded upstream). Could you modify the /docs/tutorials/advanced/tournament_results.rst file, specifically where it says:

    >>> results.initial_cooperation_count
     [9.0, 0.0, 9.0, 9.0]

could you change that to:

    >>> results.initial_cooperation_count
     [9, 0, 9, 9]

Once you've done that, commit the change and push (like we did together), this will update things here.

[Nikoleta-v3]

Ah of course 🤦‍♀️ Thank you for the clarification @drvinceknight

[marcharper]

This also happened on #1288, it may be a dask or pandas issue with int vs float types.

[drvinceknight]

Thanks for making that change @caddycarine, the tests have failed for another reason but this is not your fault. Can you modify the .github/workflows/config.yml file so that:

    - name: Check imports are sorted
       run: |
         python -m pip install isort
         python -m isort --check-only

becomes:

    - name: Check imports are sorted
       run: |
         python -m pip install "isort==4.3.21"
         python -m isort --check-only --recursive axelrod/.

Once you've done that commit and push as before :)

[marcharper]

Since all the tests pass, I think this is fine to go in as is. We'll rerun isort and black once #1288 goes in, and we'll have a script to run for all this once #1360 is done.