Added Rowsam (k58r) strategy from Axelrod's second.

[gaffney2010]

Adapted from https://github.com/Axelrod-Python/TourExec/blob/v0.3.0/src/strategies/k58r.f

In adapting:

• I called KAM distrust_points.
• I refactored NPHA in "Coop Def cycle" modes.
• I changed the K to M comparators in lines 12-15 to a comparison on K/M, which I call points_per_turn.

[marcharper]

Hi, thanks for the implementation. Do you know how to run the fingerprints for the two implementations (this and the original from axelrod/fortran)? That will help us test the equivalence, which can be tricky otherwise.

Also note that there will likely be some significant delays in review over the holidays.

[gaffney2010]

Yes, I can run them. I'm on command line only until Wednesday; will run it then.

[gaffney2010]

Bit of a delay on these; I had trouble getting everything setup.

[marcharper]

Looks good to me. Is it possible to generate fingerprints against more strategies? I think previously we've tossed in Random(p) for e.g. p in np.arange(0, 1, 0.05) as well.

[gaffney2010]

Here they are.

[drvinceknight]

Apologies for the delay in getting to this, just got back online :)

Looks great and thanks @gaffney2010 👍