min_max

multiagent/multiagent/multiAgents.py

@@ -144,27 +144,96 @@ class MultiAgentSearchAgent(Agent):
 
 class MinimaxAgent(MultiAgentSearchAgent):
     """
-      Your minimax agent (question 2)
+    Your minimax agent (question 2)
     """
-
+    def minimax(self,agentIndex,gameState,Depth):
+        if Depth >= self.depth * 2 or gameState.isWin() or gameState.isLose():
+            return self.evaluationFunction(gameState)
+        if agentIndex == 0:
+            MAX = float("-inf")
+            pacmanActions = gameState.getLegalActions(agentIndex)
+            for each_action in pacmanActions:
+                successor_state = gameState.generateSuccessor(agentIndex, each_action)
+                value = self.minimax(1, successor_state, Depth + 1)
+                MAX = max(MAX, value)
+            return MAX
+        else:
+            values = []
+            MIN = float("inf")
+            ghostActions = gameState.getLegalActions(agentIndex)
+            for each_action in ghostActions:
+                successor_state = gameState.generateSuccessor(agentIndex, each_action)
+                if agentIndex >= gameState.getNumAgents() - 1:
+                    values.append(self.minimax(0, successor_state, Depth + 1))
+                else:
+                    values.append(self.minimax(agentIndex + 1, successor_state, Depth))
+                value = min(values)
+                MIN = min(MIN, value)
+            return MIN
     def getAction(self, gameState):
         """
-          Returns the minimax action from the current gameState using self.depth
-          and self.evaluationFunction.
+        Returns the minimax action from the current gameState using self.depth
+        and self.evaluationFunction.
+
+        Here are some method calls that might be useful when implementing minimax.
+
+        gameState.getLegalActions(agentIndex):
+        Returns a list of legal actions for an agent
+        agentIndex=0 means Pacman, ghosts are >= 1
 
-          Here are some method calls that might be useful when implementing minimax.
+        gameState.generateSuccessor(agentIndex, action):
+        Returns the successor game state after an agent takes an action
 
-          gameState.getLegalActions(agentIndex):
-            Returns a list of legal actions for an agent
-            agentIndex=0 means Pacman, ghosts are >= 1
+        gameState.getNumAgents():
+        Returns the total number of agents in the game
 
-          gameState.generateSuccessor(agentIndex, action):
-            Returns the successor game state after an agent takes an action
+        gameState.isWin():
+        Returns whether or not the game state is a winning state
 
-          gameState.getNumAgents():
-            Returns the total number of agents in the game
+        gameState.isLose():
+        Returns whether or not the game state is a losing state
         """
         "*** YOUR CODE HERE ***"
+        '''
+        GhostIndex = [i for i in range(1, gameState.getNumAgents())]
+
+        def gameOver(state, d):
+            return state.isWin() or state.isLose() or d == self.depth
+        def min_value(state, d, ghost):  # minimizer
+
+            if gameOver(state, d):
+                return self.evaluationFunction(state)
+
+            Beta = 10000000000000000
+            for action in state.getLegalActions(ghost):
+                if ghost == GhostIndex[-1]:
+                    Beta = min(Beta, max_value(state.generateSuccessor(ghost, action), d + 1))
+                else:
+                    Beta = min(Beta, min_value(state.generateSuccessor(ghost, action), d, ghost + 1))
+            return Beta
+
+        def max_value(state, d):  # maximizer
+            if gameOver(state, d):
+                return self.evaluationFunction(state)
+            Alpha = -10000000000000000
+            for action in state.getLegalActions(0):
+                Alpha = max(Alpha, min_value(state.generateSuccessor(0, action), d, 1))
+            return Alpha
+        res = [(action, min_value(gameState.generateSuccessor(0, action), 0, 1)) for action in
+               gameState.getLegalActions(0)]
+        res.sort(key=lambda k: k[1])
+        return res[-1][0]
+        '''
+
+        bestaction = "stop"
+        maxvalue = float('-inf')
+        for each_action in gameState.getLegalActions(0):
+            successor_state = gameState.generateSuccessor(0, each_action)
+            value = self.minimax(agentIndex=0, gameState=successor_state, Depth=0)
+            if value is not None and maxvalue < value:
+                maxvalue = value
+                bestaction = each_action
+        return bestaction
         util.raiseNotDefined()
 
 class AlphaBetaAgent(MultiAgentSearchAgent):