back propogation

src/DecMCTS.py

@@ -194,8 +194,25 @@ class Tree:
         best_reward = float("-Inf")
         for i in range(nsims):
             # TODO
+            # SIMULATION NOT REQUIRED FOR BIGPRINT, HEURISTIC USED INSTEAD
             pass
+        avg_reward = self.reward(self.data, state)
+        self.graph.nodes["mu"] = avg_reward
+        self.graph.nodes["N"] = 1
+        
+        ### BACKPROPOGATION
+        while start_node!=1: #while not root node
+
+            start_node = self._parent(start_node)
+
+            self.graph.nodes[start_node]["mu"] = \
+                    (gamma * self.graph.nodes[start_node]["mu"] + avg_reward)\
+                    /(gamma * self.graph.nodes[start_node]["N"] + 1)
+
+            self.graph.nodes[start_node]["N"] = \
+                    gamma * self.graph.nodes[start_node]["N"] + 1
 
+        return avg_reward
 
 
     def send_comms(self):