def backFill(self, leaf, value, breadcrumbs):
lg.logger_mcts.debug('------DOING BACKFILL------')
currentPlayer = GameState.current_player_from_id(leaf.state_id)
for edge in breadcrumbs:
playerTurn = GameState.current_player_from_id(edge.inNode.state_id)
if playerTurn == currentPlayer:
direction = 1
else:
direction = -1
edge.stats['N'] = edge.stats['N'] + 1
edge.stats['W'] = edge.stats['W'] + value * direction
edge.stats['Q'] = edge.stats['W'] / edge.stats['N']
lg.logger_mcts.debug(
'updating edge with value %f for player %d... N = %d, W = %f, Q = %f',
value * direction, playerTurn, edge.stats['N'],
edge.stats['W'], edge.stats['Q'])
if lg.logger_mcts.isEnabledFor(logging.DEBUG):
lg.logger_mcts.debug(
GameState.from_id(edge.outNode.state_id,
config.GRID_SHAPE).render())
def evaluateLeaf(self, leaf, value, done, breadcrumbs):
lg.logger_mcts.debug('------EVALUATING LEAF------')
if not done:
state = GameState.from_id(leaf.state_id, config.GRID_SHAPE)
value, probs, allowedActions = self.get_preds(state)
lg.logger_mcts.debug('PREDICTED VALUE FOR %d: %f',
state.currentPlayer, value)
for idx, allowedAction in enumerate(allowedActions):
if allowedAction:
newState, _, _ = state.takeAction(idx)
if newState.id not in self.mcts.tree:
node = mc.Node(newState)
self.mcts.addNode(node)
lg.logger_mcts.debug('added node...%s...p = %f',
node.state_id, probs[idx])
else:
node = self.mcts.tree[newState.id]
lg.logger_mcts.debug('existing node...%s...',
node.state_id)
newEdge = mc.Edge(leaf, node, probs[idx], idx)
leaf.edges.append((idx, newEdge))
else:
lg.logger_mcts.debug(
'GAME VALUE FOR %d: %f',
GameState.current_player_from_id(leaf.state_id), value)
return ((value, breadcrumbs))
