stuff

MCTS.py

@@ -82,9 +82,9 @@ def moveToLeaf(self):
 
 				Q = edge.stats['Q']
 
-				lg.logger_mcts.info('action: %d (%d)... N = %d, P = %f, nu = %f, adjP = %f, W = %f, Q = %f, U = %f, Q+U = %f'
-					, action, action % 7, edge.stats['N'], np.round(edge.stats['P'],6), np.round(nu[idx],6), ((1-epsilon) * edge.stats['P'] + epsilon * nu[idx] )
-					, np.round(edge.stats['W'],6), np.round(Q,6), np.round(U,6), np.round(Q+U,6))
+				# lg.logger_mcts.info('action: %d (%d)... N = %d, P = %f, nu = %f, adjP = %f, W = %f, Q = %f, U = %f, Q+U = %f'
+				# 	, action, action % 7, edge.stats['N'], np.round(edge.stats['P'],6), np.round(nu[idx],6), ((1-epsilon) * edge.stats['P'] + epsilon * nu[idx] )
+				# 	, np.round(edge.stats['W'],6), np.round(Q,6), np.round(U,6), np.round(Q+U,6))
 
 				if Q + U > maxQU:
 					maxQU = Q + U

agent.py

@@ -23,7 +23,8 @@ def __init__(self, name, state_size, action_size):
 		self.action_size = action_size
 
 	def act(self, state, tau):
-		action = input('Enter your chosen action: ')
+		print(tau)
+		action = int(input('Enter your chosen action: '))
 		pi = np.zeros(self.action_size)
 		pi[action] = 1
 		value = None
@@ -91,6 +92,8 @@ def act(self, state, tau):
 		####pick the action
 		action, value = self.chooseAction(pi, values, tau)
 
+		print("action:"+str(action))
+
 		nextState, _, _ = state.takeAction(action)
 
 		NN_value = -self.get_preds(nextState)[0]