Moran plot #1122
Moran plot #1122
Conversation
| @@ -385,6 +386,28 @@ def __len__(self) -> int: | |||
| """ | |||
| return len(self.populations) | |||
|
|
|||
| def populations_plot(self, ax=None): | |||
There was a problem hiding this comment.
Could we just have a docstring?
Something like (but please modify/improve):
Plot the population distribution over iterations of the Moran process.
Parameters
----------------
ax: a matplotlib axis
Allows to plot this on a given matplotlib axis. Default is None.
Returns
-----------
A matplotlibt axis object
| >>> import matplotlib.pyplot as plt | ||
| >>> axl.seed(15) # for reproducible example | ||
| >>> N = 10 | ||
| >>> players = [random.choice(axl.basic_strategies)() for i in range(N)] |
There was a problem hiding this comment.
Could we hard code a list of strategies please. Otherwise, if and when basic_strategies changes (with new strategies being added) this would not necessarily give the same plot.
| >>> mp.winning_strategy_name | ||
| 'Tit For Tat' | ||
| >>> ax = mp.populations_plot() | ||
| >>> plt.show() #doctest: +SKIP |
There was a problem hiding this comment.
Actually the test hangs and all the windows stay open otherwise, something to do with plt not exiting perhaps because the plot object isn't gc'd.
There was a problem hiding this comment.
gc'd (curious)?
No problem to skip.
axelrod/tests/unit/test_moran.py
Outdated
| # Test that can plot on a given matplotlib axes | ||
| axelrod.seed(15) | ||
| N = 5 | ||
| players = [random.choice(axelrod.basic_strategies)() for _ in range(N)] |
There was a problem hiding this comment.
Same comment about the basic_strategies would this potentially change the axes in lines 345, 346 if and when new strategies come in?
For the purpose of the test let's just use demo_strategies (or any other hard coded list: perhaps just a cooperator and a defector?)
axelrod/tests/unit/test_moran.py
Outdated
| axelrod.seed(15) | ||
| N = 5 | ||
| players = [random.choice(axelrod.basic_strategies)() for _ in range(N)] | ||
| mp = axelrod.MoranProcess(players=players, turns=200) |
There was a problem hiding this comment.
For the purpose of the test can we reduce turns? (for run time?)
There was a problem hiding this comment.
Sure but there's a much longer test in the moran_test.py file.
There was a problem hiding this comment.
Feel free to reduce that one too :)
axelrod/tests/unit/test_moran.py
Outdated
| mp.play() | ||
| fig, axarr = plt.subplots(2, 2) | ||
| ax = axarr[1, 0] | ||
| mp.populations_plot(ax=ax) |
There was a problem hiding this comment.
Can we also have a test where we just call mp.populations_plot() (to test the default runs without a problem).
There was a problem hiding this comment.
(This is where the coverage is missing)
|
FYI all the tests pass locally, looks like a timeout issue. |
Is this rebased on to master with the stuff I did on #1121 ? |
axelrod/tests/unit/test_moran.py
Outdated
| self.assertEqual(ax.get_xlim(), (-0.8, 16.8)) | ||
| self.assertEqual(ax.get_ylim(), (0, 5.25)) | ||
|
|
||
| # Run with a give axis |
There was a problem hiding this comment.
# Run without a given axis (I guess)
axelrod/tests/unit/test_moran.py
Outdated
| self.assertEqual(ax.get_ylim(), (0, 5.25)) | ||
|
|
||
| # Run with a give axis | ||
| mp.populations_plot() |
There was a problem hiding this comment.
Can we test something.
Perhaps (to avoid adding to the previous plot):
plt.figure()
self.assertIsInstance(mp.populations_plot(), matplotlib.pyplot.Figure) # I'm guessing?
| >>> import matplotlib.pyplot as plt | ||
| >>> axl.seed(15) # for reproducible example | ||
| >>> N = 10 | ||
| >>> players = [random.choice(axl.demo_strategies)() for i in range(N)] |
There was a problem hiding this comment.
Thanks for changing this, the plot file now needs to be updated to run with these strategies (eg Suspicious Tit For Tat is not in demo_strategies). Another alternative is to just hard code the names of the players in that plot file (instead of random.choice(axl.demo_strategies)) either way they need to match.
There was a problem hiding this comment.
I think I would prefer the hard coded approach as there's less code to (potentially) understand but your call. As long as they match up.
It's looking like it's not, the PR run worked fine (and the log shows the timeout wait call) but the push didn't. If you rebase on to |
Add a plot for the Moran process. Closes #1048 .