'''
DoubleQAgentClass.py: Class for an RL Agent acting according to Double Q Learning from:
Hasselt, H. V. (2010). Double Q-learning.
In Advances in Neural Information Processing Systems (pp. 2613-2621).
Author: David Abel
'''
# Python imports.
import random
from collections import defaultdict
# Other imports
from simple_rl.agents.QLearningAgentClass import QLearningAgent
from simple_rl.agents.AgentClass import Agent
[docs]class DoubleQAgent(QLearningAgent):
''' Class for an agent using Double Q Learning. '''
def __init__(self, actions, name="Double-Q", alpha=0.05, gamma=0.99, epsilon=0.1, explore="uniform", anneal=False):
'''
Args:
actions (list): Contains strings denoting the actions.
name (str): Denotes the name of the agent.
alpha (float): Learning rate.
gamma (float): Discount factor.
epsilon (float): Exploration term.
explore (str): One of {softmax, uniform}. Denotes explore policy.
'''
QLearningAgent.__init__(self, actions, name=name, alpha=alpha, gamma=gamma, epsilon=epsilon, explore=explore, anneal=anneal)
# Make two q functions.
self.q_funcs = {"A":defaultdict(lambda : defaultdict(lambda: self.default_q)), \
"B":defaultdict(lambda : defaultdict(lambda: self.default_q))}
[docs] def act(self, state, reward):
'''
Args:
state (State)
reward (float)
Summary:
The central method called during each time step.
Retrieves the action according to the current policy
and performs updates.
'''
self.update(self.prev_state, self.prev_action, reward, state)
if self.explore == "softmax":
# Softmax exploration
action = self.soft_max_policy(state)
else:
# Uniform exploration
action = self.epsilon_greedy_q_policy(state)
self.prev_state = state
self.prev_action = action
self.step_number += 1
# Anneal params.
if self.anneal:
self._anneal()
return action
[docs] def update(self, state, action, reward, next_state):
'''
Args:
state (State)
action (str)
reward (float)
next_state (State)
Summary:
Updates the internal Q Function according to the Double Q update:
'''
# If this is the first state, just return.
if state is None:
self.prev_state = next_state
return
# Randomly choose either "A" or "B".
which_q_func = "A" if bool(random.getrandbits(1)) else "B"
other_q_func = "B" if which_q_func is "A" else "A"
# Update the Q Function.
# Get max q action of the chosen Q func.
max_q_action = self.get_max_q_action(next_state, q_func_id=which_q_func)
prev_q_val = self.get_q_value(state, action, q_func_id=which_q_func)
# Update
self.q_funcs[which_q_func][state][action] = (1 - self.alpha) * prev_q_val + self.alpha * (reward + self.gamma * self.get_q_value(next_state, max_q_action, q_func_id=other_q_func))
[docs] def get_max_q_action(self, state, q_func_id=None):
'''
Args:
state (State)
q_func_id (str): either "A" or "B"
Returns:
(str): denoting the action with the max q value in the given @state.
'''
return self._compute_max_qval_action_pair(state, q_func_id)[1]
[docs] def get_max_q_value(self, state, q_func_id=None):
'''
Args:
state (State)
q_func_id (str): either "A" or "B"
Returns:
(float): denoting the max q value in the given @state.
'''
return self._compute_max_qval_action_pair(state, q_func_id)[0]
def _compute_max_qval_action_pair(self, state, q_func_id=None):
'''
Args:
state (State)
q_func_id (str): either "A", "B", or None. If None, computes avg of A and B.
Returns:
(tuple) --> (float, str): where the float is the Qval, str is the action.
'''
# Grab random initial action in case all equal
best_action = random.choice(self.actions)
max_q_val = float("-inf")
shuffled_action_list = self.actions[:]
random.shuffle(shuffled_action_list)
# Find best action (action w/ current max predicted Q value)
for action in shuffled_action_list:
q_s_a = self.get_q_value(state, action, q_func_id)
if q_s_a > max_q_val:
max_q_val = q_s_a
best_action = action
return max_q_val, best_action
[docs] def get_q_value(self, state, action, q_func_id=None):
'''
Args:
state (State)
action (str)
q_func_id (str): either "A", "B", or defaults to taking the average.
Returns:
(float): denoting the q value of the (@state, @action) pair relative to
the specified q function.
'''
if q_func_id is None:
return self.get_avg_q_value(state, action)
else:
return self.q_funcs[q_func_id][state][action]
[docs] def reset(self):
self.step_number = 0
self.episode_number = 0
self.q_funcs = {"A":defaultdict(lambda : defaultdict(lambda: self.default_q)), \
"B":defaultdict(lambda : defaultdict(lambda: self.default_q))}
Agent.reset(self)
# ---- DOUBLE Q NEW ----
[docs] def get_avg_q_value(self, state, action):
'''
Args:
state (State)
action (str)
Returns:
(float): denoting the avg. q value of the (@state, @action) pair.
'''
return (self.q_funcs["A"][state][action] + self.q_funcs["B"][state][action]) / 2.0