Merge pull request #1091 from dmanc/master

Michaelos strategy

Author: meatballs

axelrod/strategies/_strategies.py

@@ -76,8 +76,7 @@
     TitForTat, TitFor2Tats, TwoTitsForTat, Bully, SneakyTitForTat,
     SuspiciousTitForTat, AntiTitForTat, HardTitForTat, HardTitFor2Tats,
     OmegaTFT, Gradual, ContriteTitForTat, AdaptiveTitForTat,
-    SpitefulTitForTat, SlowTitForTwoTats2, Alexei, EugineNier,
-    DynamicTwoTitsForTat, NTitsForMTats)
+    SpitefulTitForTat, SlowTitForTwoTats2, Alexei, EugineNier, DynamicTwoTitsForTat, NTitsForMTats, Michaelos)
 from .verybad import VeryBad
 from .worse_and_worse import (WorseAndWorse, KnowledgeableWorseAndWorse,
                               WorseAndWorse2, WorseAndWorse3)
@@ -191,6 +190,7 @@
     MathConstantHunter,
     NaiveProber,
     MEM2,
+    Michaelos,
     MindBender,
     MindController,
     MindReader,

axelrod/strategies/titfortat.py

@@ -778,3 +778,52 @@ def strategy(self, opponent: Player) -> Action:
     def reset(self):
         super().reset()
         self.retaliate_count = 0
+
+
+@FinalTransformer((D,), name_prefix=None)
+class Michaelos(Player):
+    """
+    Plays similar to Tit-for-Tat with two exceptions:
+    1) Defect on last turn.
+    2) After own defection and opponent's cooperation, 50 percent of the time,
+    cooperate. The other 50 percent of the time, always defect for the rest of
+    the game.
+
+    Names:
+
+    - Michaelos: [LessWrong2011]_
+    """
+
+    name = 'Michaelos'
+    classifier = {
+        'memory_depth': float('inf'),
+        'stochastic': True,
+        'makes_use_of': {'length'},
+        'long_run_time': False,
+        'inspects_source': False,
+        'manipulates_source': False,
+        'manipulates_state': False
+    }
+
+    def __init__(self):
+        super().__init__()
+        self.is_defector = False
+
+    def strategy(self, opponent: Player) -> Action:
+        if not self.history:
+            return C
+        if self.is_defector:
+            return D
+        if self.history[-1] == D and opponent.history[-1] == C:
+            decision = random_choice()
+            if decision == C:
+                return C
+            else:
+                self.is_defector = True
+                return D
+
+        return opponent.history[-1]
+
+    def reset(self):
+        super().reset()
+        self.is_defector = False

axelrod/tests/strategies/test_titfortat.py

@@ -717,3 +717,54 @@ def test_varying_memory_depth(self):
         self.assertEqual(self.player(1, 1).classifier['memory_depth'], 1)
         self.assertEqual(self.player(0, 3).classifier['memory_depth'], 3)
         self.assertEqual(self.player(5, 3).classifier['memory_depth'], 5)
+
+
+class TestMichaelos(TestPlayer):
+    """
+    Tests for the Michaelos strategy
+    """
+
+    name = "Michaelos: (D,)"
+    player = axelrod.Michaelos
+    expected_classifier = {
+        'memory_depth': float('inf'),
+        'stochastic': True,
+        'makes_use_of': {'length'},
+        'inspects_source': False,
+        'manipulates_source': False,
+        'manipulates_state': False
+    }
+
+    def test_strategy(self):
+
+        actions = [(C, C), (C, C), (C, C), (D, C)]
+        self.versus_test(axelrod.Cooperator(), expected_actions=actions,
+                         attrs={"is_defector": False}, seed=2)
+
+        actions = [(C, C), (C, C), (C, C), (C, C)]
+        self.versus_test(axelrod.Cooperator(), expected_actions=actions,
+                         attrs={"is_defector": False},
+                         match_attributes={"length": -1}, seed=2)
+
+        actions = [(C, D), (D, D), (D, D), (D, D)]
+        self.versus_test(axelrod.Defector(), expected_actions=actions,
+                         attrs={"is_defector": False}, seed=2)
+
+        actions = [(C, D), (D, D), (D, D), (D, D)]
+        self.versus_test(axelrod.Defector(), expected_actions=actions,
+                         attrs={"is_defector": False},
+                         match_attributes={"length": -1}, seed=2)
+
+        # Chance of becoming a defector is 50% after (D, C) occurs.
+        actions = [(C, C), (C, D), (D, C), (C, D), (D, C)]
+        self.versus_test(axelrod.Alternator(), expected_actions=actions,
+                         attrs={"is_defector": False}, seed=3)
+
+        actions = [(C, C), (C, D), (D, C), (D, D), (D, C), (D, D), (D, C)]
+        self.versus_test(axelrod.Alternator(), expected_actions=actions,
+                         attrs={"is_defector": True}, seed=2)
+
+        actions = [(C, C), (C, D), (D, C), (C, D), (D, C), (D, D), (D, C)]
+        self.versus_test(axelrod.Alternator(), expected_actions=actions,
+                         attrs={"is_defector": True},
+                         match_attributes={"length": -1}, seed=1)

docs/tutorials/advanced/classification_of_strategies.rst

@@ -47,7 +47,7 @@ strategies::
     ... }
     >>> strategies = axl.filtered_strategies(filterset)
     >>> len(strategies)
-    71
+    72
 
 Or, to find out how many strategies only use 1 turn worth of memory to
 make a decision::
@@ -80,7 +80,7 @@ length of each match of the tournament::
     ... }
     >>> strategies = axl.filtered_strategies(filterset)
     >>> len(strategies)
-    28
+    29
 
 Note that in the filterset dictionary, the value for the 'makes_use_of' key
 must be a list. Here is how we might identify the number of strategies that use