dqn_agent.py

import numpy as np
import random
from collections import namedtuple, deque

import torch
import torch.nn.functional as F
import torch.optim as optim
import qnetwork


device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

    
class Agent():
    """Interacts with and learns from the environment."""

    @staticmethod
    def default_parameter():
        return {
             'state_size': 37, 
             'action_size': 4, 
             'hidden_layers': [64,64],
             'drop_p': 0.0,
             'seed': 0,
             'GAMMA': 0.99,
             'LR': 0.0005,
             'UPDATE_EVERY': 25,
             'BUFFER_SIZE': 100000,
             'BATCH_SIZE': 64,
             'TAU': 0.001}

    def __init__(self, parameter):
        """Initialize an Agent object.
        
        Params
        ======
            parameters            
        """
        self.state_size = parameter['state_size']           # state_size (int): dimension of each state
        self.action_size = parameter['action_size']         # action_size (int): dimension of each action  
        self.seed = random.seed(parameter['seed'])          # random seed
        self.GAMMA = parameter['GAMMA']                     # discount factor
        self.LR = parameter['LR']                           # learning rate
        self.UPDATE_EVERY = parameter['UPDATE_EVERY']       # how often to update the network
        self.BUFFER_SIZE = parameter['BUFFER_SIZE']         # replay buffer size
        self.BATCH_SIZE = parameter['BATCH_SIZE']           # minibatch size
        self.TAU = parameter['TAU']                         # for soft update of target parameters

        # Q-Network
        self.qnetwork_local =  qnetwork.Network(self.state_size, self.action_size, parameter['hidden_layers'], parameter['drop_p']).to(device)
        self.qnetwork_target = qnetwork.Network(self.state_size, self.action_size, parameter['hidden_layers'], parameter['drop_p']).to(device)
        self.optimizer = optim.Adam(self.qnetwork_local.parameters(), lr=self.LR)

        # Replay memory
        self.memory = ReplayBuffer(self.action_size, self.BUFFER_SIZE, self.BATCH_SIZE, self.seed)
        # Initialize time step (for updating every UPDATE_EVERY steps)
        self.t_step = 0
        
        print('\nAgent Device: {}'.format(device))
    
    def step(self, state, action, reward, next_state, done):
        # Save experience in replay memory
        self.memory.add(state, action, reward, next_state, done)
        
        # Learn every UPDATE_EVERY time steps.
        self.t_step = (self.t_step + 1) % self.UPDATE_EVERY
        if self.t_step == 0:
            # If enough samples are available in memory, get random subset and learn
            if len(self.memory) > self.BATCH_SIZE:
                experiences = self.memory.sample()
                self.learn(experiences, self.GAMMA)

    def act(self, state, eps=0.):
        """Returns actions for given state as per current policy.
        
        Params
        ======
            state (array_like): current state
            eps (float): epsilon, for epsilon-greedy action selection
        """
        state = torch.from_numpy(state).float().unsqueeze(0).to(device)
        self.qnetwork_local.eval()
        with torch.no_grad():
            action_values = self.qnetwork_local(state)
        self.qnetwork_local.train()

        # Epsilon-greedy action selection
        if random.random() > eps:
            return np.argmax(action_values.cpu().data.numpy())
        else:
            return random.choice(np.arange(self.action_size))

    def learn(self, experiences, gamma):
        """Update value parameters using given batch of experience tuples.

        Params
        ======
            experiences (Tuple[torch.Variable]): tuple of (s, a, r, s', done) tuples 
            gamma (float): discount factor
        """
        states, actions, rewards, next_states, dones = experiences

        # Get max predicted Q values (for next states) from target model
        Q_targets_next = self.qnetwork_target(next_states).detach().max(1)[0].unsqueeze(1)
        # Compute Q targets for current states 
        Q_targets = rewards + (gamma * Q_targets_next * (1 - dones))

        # Get expected Q values from local model
        Q_expected = self.qnetwork_local(states).gather(1, actions)

        # Compute loss
        loss = F.mse_loss(Q_expected, Q_targets)
        # Minimize the loss
        self.optimizer.zero_grad()
        loss.backward()
        self.optimizer.step()

        # ------------------- update target network ------------------- #
        self.soft_update(self.qnetwork_local, self.qnetwork_target, self.TAU)                     

    def soft_update(self, local_model, target_model, tau):
        """Soft update model parameters.
        θ_target = τ*θ_local + (1 - τ)*θ_target

        Params
        ======
            local_model (PyTorch model): weights will be copied from
            target_model (PyTorch model): weights will be copied to
            tau (float): interpolation parameter 
        """
        for target_param, local_param in zip(target_model.parameters(), local_model.parameters()):
            target_param.data.copy_(tau*local_param.data + (1.0-tau)*target_param.data)

    def save_checkpoint(self, file):
        checkpoint = {'input_size': self.qnetwork_local.input_size,
              'output_size': self.qnetwork_local.output_size,
              'hidden_layers': [each.out_features for each in self.qnetwork_local.hidden_layers],
              'drop_p': self.qnetwork_local.drop_p,
              'state_dict': self.qnetwork_local.state_dict()}
        torch.save(checkpoint, file)
        
    def load_checkpoint(self, file):
        checkpoint = torch.load(file)
        self.qnetwork_local = qnetwork.Network(checkpoint['input_size'],
                                 checkpoint['output_size'],
                                 checkpoint['hidden_layers'],
                                 checkpoint['drop_p']).to(device)
        self.qnetwork_local.load_state_dict(checkpoint['state_dict'])
        self.qnetwork_target = qnetwork.Network(checkpoint['input_size'],
                                 checkpoint['output_size'],
                                 checkpoint['hidden_layers'],
                                 checkpoint['drop_p']).to(device)
        self.qnetwork_target.load_state_dict(checkpoint['state_dict'])

class ReplayBuffer:
    """Fixed-size buffer to store experience tuples."""

    def __init__(self, action_size, buffer_size, batch_size, seed):
        """Initialize a ReplayBuffer object.

        Params
        ======
            action_size (int): dimension of each action
            buffer_size (int): maximum size of buffer
            batch_size (int): size of each training batch
            seed (int): random seed
        """
        self.action_size = action_size
        self.memory = deque(maxlen=buffer_size)  
        self.batch_size = batch_size
        self.experience = namedtuple("Experience", field_names=["state", "action", "reward", "next_state", "done"])
        self.seed = random.seed(seed)
    
    def add(self, state, action, reward, next_state, done):
        """Add a new experience to memory."""
        e = self.experience(state, action, reward, next_state, done)
        self.memory.append(e)
    
    def sample(self):
        """Randomly sample a batch of experiences from memory."""
        experiences = random.sample(self.memory, k=self.batch_size)

        states = torch.from_numpy(np.vstack([e.state for e in experiences if e is not None])).float().to(device)
        actions = torch.from_numpy(np.vstack([e.action for e in experiences if e is not None])).long().to(device)
        rewards = torch.from_numpy(np.vstack([e.reward for e in experiences if e is not None])).float().to(device)
        next_states = torch.from_numpy(np.vstack([e.next_state for e in experiences if e is not None])).float().to(device)
        dones = torch.from_numpy(np.vstack([e.done for e in experiences if e is not None]).astype(np.uint8)).float().to(device)
  
        return (states, actions, rewards, next_states, dones)

    def __len__(self):
        """Return the current size of internal memory."""
        return len(self.memory)