Add new gradual and discussion of tournament #1299
Conversation
Added more to description as well. I've also added the https://github.com/cristal-smac/ipd repo to our bibliography. (Using 2018 which was the date of their first commit).
I've added the longer tests from the discussion at #1294
| Beaufils et al.'s tournament (1997) | ||
| ----------------------------------- | ||
|
|
||
| In 1997, [Beaufils1997]_ presented used a tournament to describe a new strategy |
|
|
||
| if len(self.history) == 0: | ||
| return C | ||
|
|
There was a problem hiding this comment.
niggle: too many empty lines (black might need running here)
There was a problem hiding this comment.
b7ff5c6 ran black on the file
|
No idea why the travis build is not appearing, it is happening here: https://travis-ci.org/github/Axelrod-Python/Axelrod/jobs/666857590?utm_medium=notification&utm_source=github_status |
I suggest not worrying about it (we can still make sure the tests pass of course by looking directly at travis) as we will be moving to GitHub actions (see #1297) |
|
Here's the link to the various branch builds: https://travis-ci.org/github/Axelrod-Python/Axelrod/branches?utm_medium=notification&utm_source=github_status |
| @@ -466,6 +689,9 @@ def test_output_from_literature(self): | |||
| Dilemma" Proc. Artif. Life 1996 | |||
|
|
|||
| This test just ensures that the strategy is as was originally defined. | |||
|
|
|||
| See https://github.com/Axelrod-Python/Axelrod/issues/1294 for another | |||
| discussion of this. | |||
There was a problem hiding this comment.
Should this be a test similar to test_specific_set_of_results for Gradual? Run the whole tournament and compare the scores?
There was a problem hiding this comment.
Note that this will just repeat the doctest. (I'm fine with that.)
There was a problem hiding this comment.
Sure. It already does that but only for two strategies right?
We are testing the output of the paper so I suggest we include all of it, and it will follow the same format as the tests for Gradual.
Instead of just checking two strategies, check the entire set of reported results in the paper.
Co-Authored-By: Nikoleta Glynatsi <GlynatsiNE@cardiff.ac.uk>
|
I think this is good to be merged now? |
Adds the new Gradual see discussion at #1294
Note I've gone with:
OriginalGradual(the version that was in the library and that was used in the 1997 paper)Gradual(the one the authors apparently meant to use).I've decided to not include
mistrustor at least the broken version of mistrust that the authors used in #1294 that's needed to reproduce the results suggested in #1294, the authors meant for mistrust to beSuspicious Tit For Tatwhich we have in the library (and this name is already noted there).Closes #1294