class PrioritizedReplayBuffer(ReplayBuffer):
def __init__(self, size, alpha):
"""Create Prioritized Replay buffer.
Parameters
----------
size: int
Max number of transitions to store in the buffer. When the buffer
overflows the old memories are dropped.
alpha: float
how much prioritization is used
(0 - no prioritization, 1 - full prioritization)
See Also
--------
ReplayBuffer.__init__
"""
super(PrioritizedReplayBuffer, self).__init__(size)
assert alpha >= 0
self._alpha = alpha
self.it_capacity = 1
while self.it_capacity < size*2: # We use double the soft capacity of the PER for the segment trees to allow for any overflow over the soft capacity limit before samples are removed
self.it_capacity *= 2
self._it_sum = SumSegmentTree(self.it_capacity)
self._it_min = MinSegmentTree(self.it_capacity)
self._max_priority = 1.0
def add(self, *args, **kwargs):
idx = self._next_idx
assert idx < self.it_capacity, "Number of samples in replay memory exceeds capacity of segment trees. Please increase capacity of segment trees or increase the frequency at which samples are removed from the replay memory"
super().add(*args, **kwargs)
self._it_sum[idx] = self._max_priority ** self._alpha
self._it_min[idx] = self._max_priority ** self._alpha
def remove(self, num_samples):
super().remove(num_samples)
self._it_sum.remove_items(num_samples)
self._it_min.remove_items(num_samples)
def _sample_proportional(self, batch_size):
res = []
p_total = self._it_sum.sum(0, len(self._storage) - 1)
every_range_len = p_total / batch_size
for i in range(batch_size):
mass = random.random() * every_range_len + i * every_range_len
idx = self._it_sum.find_prefixsum_idx(mass)
res.append(idx)
return res
def sample(self, batch_size, beta):
"""Sample a batch of experiences.
compared to ReplayBuffer.sample
it also returns importance weights and idxes
of sampled experiences.
Parameters
----------
batch_size: int
How many transitions to sample.
beta: float
To what degree to use importance weights
(0 - no corrections, 1 - full correction)
Returns
-------
obs_batch: np.array
batch of observations
act_batch: np.array
batch of actions executed given obs_batch
rew_batch: np.array
rewards received as results of executing act_batch
next_obs_batch: np.array
next set of observations seen after executing act_batch
done_mask: np.array
done_mask[i] = 1 if executing act_batch[i] resulted in
the end of an episode and 0 otherwise.
gammas: np.array
product of gammas for N-step returns
weights: np.array
Array of shape (batch_size,) and dtype np.float32
denoting importance weight of each sampled transition
idxes: np.array
Array of shape (batch_size,) and dtype np.int32
idexes in buffer of sampled experiences
"""
assert beta > 0
idxes = self._sample_proportional(batch_size)
weights = []
p_min = self._it_min.min() / self._it_sum.sum()
max_weight = (p_min * len(self._storage)) ** (-beta)
for idx in idxes:
p_sample = self._it_sum[idx] / self._it_sum.sum()
weight = (p_sample * len(self._storage)) ** (-beta)
weights.append(weight / max_weight)
weights = np.array(weights)
encoded_sample = self._encode_sample(idxes)
return tuple(list(encoded_sample) + [weights, idxes])
def update_priorities(self, idxes, priorities):
"""Update priorities of sampled transitions.
sets priority of transition at index idxes[i] in buffer
to priorities[i].
Parameters
----------
idxes: [int]
List of idxes of sampled transitions
priorities: [float]
List of updated priorities corresponding to
transitions at the sampled idxes denoted by
variable `idxes`.
"""
assert len(idxes) == len(priorities)
for idx, priority in zip(idxes, priorities):
assert priority > 0
assert 0 <= idx < len(self._storage)
self._it_sum[idx] = priority ** self._alpha
self._it_min[idx] = priority ** self._alpha
self._max_priority = max(self._max_priority, priority)
class PrioritizedReplayBuffer(ReplayBuffer):
def __init__(self, max_size, alpha):
"""Create Prioritized Replay buffer.
Parameters
----------
max_size: int
Max number of transitions to store in the buffer.
When the buffer overflows the old memories are dropped.
alpha: float
how much prioritization is used (0: no prioritization, 1: full prioritization)
See Also
--------
ReplayBuffer.__init__
"""
super(PrioritizedReplayBuffer, self).__init__(max_size)
assert alpha >= 0
self._alpha = alpha
it_capacity = 1
while it_capacity < max_size:
it_capacity *= 2
self._it_sum = SumSegmentTree(it_capacity)
self._it_min = MinSegmentTree(it_capacity)
self._max_priority = 1.0
def add(self, state, action, next_state, reward, done):
"""See ReplayBuffer.add"""
idx = self.ptr
super().add(state, action, next_state, reward, done)
self._it_sum[idx] = self._max_priority**self._alpha
self._it_min[idx] = self._max_priority**self._alpha
def _sample_proportional(self, batch_size):
res = []
p_total = self._it_sum.sum(0, len(self._storage) - 1)
every_range_len = p_total / batch_size
for i in range(batch_size):
mass = random.random() * every_range_len + i * every_range_len
idx = self._it_sum.find_prefixsum_idx(mass)
res.append(idx)
return res
def sample(self, batch_size, beta):
"""Sample a batch of experiences.
compared to ReplayBuffer.sample
it also returns importance weights and idxes of sampled experiences.
Parameters
----------
batch_size: int
How many transitions to sample.
beta: float
To what degree to use importance weights (0: no corrections, 1: full correction)
Returns
-------
obs_batch: np.array
batch of observations
act_batch: np.array
batch of actions executed given obs_batch
rew_batch: np.array
rewards received as results of executing act_batch
next_obs_batch: np.array
next set of observations seen after executing act_batch
done_mask: np.array
done_mask[i] = 1 if executing act_batch[i] resulted in
the end of an episode and 0 otherwise.
weights: np.array
Array of shape (batch_size,) and dtype np.float32
denoting importance weight of each sampled transition
idxes: np.array
Array of shape (batch_size,) and dtype np.int32
idexes in buffer of sampled experiences
"""
assert beta > 0
idxes = self._sample_proportional(batch_size)
weights = []
p_min = self._it_min.min() / self._it_sum.sum()
max_weight = (p_min * len(self._storage))**(-beta)
for idx in idxes:
p_sample = self._it_sum[idx] / self._it_sum.sum()
weight = (p_sample * len(self._storage))**(-beta)
weights.append(weight / max_weight)
weights = np.array(weights)
encoded_sample = self._encode_sample(idxes)
return tuple(
list(encoded_sample) + [
torch.as_tensor(
weights, device=self.device, dtype=torch.float32), idxes
])
def update_priorities(self, idxes, priorities):
"""Update priorities of sampled transitions.
Sets priority of transition at index idxes[i] in buffer to priorities[i].
Parameters
----------
idxes: [int]
List of idxes of sampled transitions
priorities: [float]
List of updated priorities corresponding to transitions at the sampled idxes denoted
by variable `idxes`.
"""
assert len(idxes) == len(priorities)
for idx, priority in zip(idxes, priorities):
assert priority > 0
assert 0 <= idx < len(self._storage)
self._it_sum[idx] = priority**self._alpha # update priority
self._it_min[idx] = priority**self._alpha
self._max_priority = max(self._max_priority, priority)
class PrioritizedReplayBuffer(ReplayBuffer):
def __init__(self, args, buffer_id):
"""Create Prioritized Replay buffer.
Parameters
----------
size: int
Max number of transitions to store in the buffer. When the buffer
overflows the old memories are dropped.
alpha: float
how much prioritization is used
(0 - no prioritization, 1 - full prioritization)
beta: float
To what degree to use importance weights
(0 - no corrections, 1 - full correction)
See Also
--------
ReplayBuffer.__init__
"""
super(PrioritizedReplayBuffer, self).__init__(args, buffer_id)
assert self.args.alpha > 0
self._alpha = args.replay_alpha
self._beta = args.replay_beta
it_capacity = 1
while it_capacity < self.args.size:
it_capacity *= 2
self._it_sum = SumSegmentTree(it_capacity)
self._it_min = MinSegmentTree(it_capacity)
self._max_priority = 1.0
def add(self, obs_t, action, reward, obs_tp1, done, weight):
"""See ReplayBuffer.store_effect"""
idx = self._next_idx
super(PrioritizedReplayBuffer, self).add(obs_t, action, reward,
obs_tp1, done, weight)
if weight is None:
weight = self._max_priority
self._it_sum[idx] = weight**self._alpha
self._it_min[idx] = weight**self._alpha
def _sample_proportional(self, batch_size):
res = []
for _ in range(batch_size):
mass = random.random() * self._it_sum.sum(0, len(self._storage))
idx = self._it_sum.find_prefixsum_idx(mass)
res.append(idx)
return np.array(res, dtype=np.int32)
def sample_idxes(self, batch_size):
return self._sample_proportional(batch_size)
def sample_with_weights_and_idxes(self, idxes):
weights = []
p_min = self._it_min.min() / self._it_sum.sum()
max_weight = (p_min * len(self._storage))**(-self._beta)
for idx in idxes:
p_sample = self._it_sum[idx] / self._it_sum.sum()
weight = (p_sample * len(self._storage))**(-self._beta)
weights.append(weight / max_weight)
weights = np.array(weights)
encoded_sample = self._encode_sample(idxes)
return list(encoded_sample) + [weights, idxes]
def sample(self, batch_size):
idxes = self.sample_idxes(batch_size)
return self.sample_with_weights_and_idxes(idxes)
def update_priorities(self, idxes, priorities):
"""Update priorities of sampled transitions.
sets priority of transition at index idxes[i] in buffer
to priorities[i].
Parameters
----------
idxes: [int]
List of idxes of sampled transitions
priorities: [float]
List of updated priorities corresponding to
transitions at the sampled idxes denoted by
variable `idxes`.
"""
assert len(idxes) == len(priorities)
for idx, priority in zip(idxes, priorities):
assert priority > 0
assert 0 <= idx < len(self._storage)
delta = priority**self._alpha - self._it_sum[idx]
self._it_sum[idx] = priority**self._alpha
self._it_min[idx] = priority**self._alpha
self._max_priority = max(self._max_priority, priority)
class PrioritizedReplayBufferTorch(ReplayBufferTorch):
def __init__(self, size, alpha, device):
#print(self.__mro__)
super().__init__(size, device)
assert alpha >= 0
self._alpha = alpha
it_capacity = 2
while it_capacity < size:
it_capacity *= 2
self._it_sum = SumSegmentTree(it_capacity)
self._it_min = MinSegmentTree(it_capacity)
self._max_priority = 1.0
def add(self, obs_t, action, reward, obs_tp1, done):
idx = self._next_idx
data = (obs_t, action, reward, obs_tp1, done)
if self._next_idx >= len(self._storage):
self._storage.append(data)
else:
self._storage[self._next_idx] = data
self._next_idx = (self._next_idx + 1) % self._maxsize
self._it_sum[idx] = self._max_priority**self._alpha
self._it_min[idx] = self._max_priority**self._alpha
def _sample_proportional(self, batch_size):
res = []
p_total = self._it_sum.sum(0, len(self._storage) - 1)
every_range_len = p_total / batch_size
for i in range(batch_size):
mass = random.random() * every_range_len + i * every_range_len
idx = self._it_sum.find_prefixsum_idx(mass)
res.append(idx)
return res
def sample(self, batch_size, beta):
assert beta > 0
idxes = self._sample_proportional(batch_size)
weights = []
p_min = self._it_min.min() / self._it_sum.sum()
max_weight = (p_min * len(self._storage))**(-beta)
for idx in idxes:
p_sample = self._it_sum[idx] / self._it_sum.sum()
weight = (p_sample * len(self._storage))**(-beta)
weights.append(weight / max_weight)
weights = torch.tensor(weights,
dtype=torch.float32,
device=self.device)
encoded_sample = self._encode_sample(idxes)
return tuple(list(encoded_sample) + [weights, idxes])
def update_priorities(self, idxes, priorities):
assert len(idxes) == len(priorities)
assert all(0 <= x < len(self._storage) for x in idxes)
assert (priorities > 0).all()
self._max_priority = max(self._max_priority, max(priorities))
for idx, priority in zip(idxes, priorities):
#assert priority > 0
#assert 0 <= idx < len(self._storage)
#print(priority)
self._it_sum[idx] = priority**self._alpha
self._it_min[idx] = priority**self._alpha
class PrioritizedReplayBuffer(ReplayBuffer):
def __init__(self,
size,
frame_history_len,
alpha,
num_branches,
non_pixel_dimension,
add_non_pixel=False):
"""
----------
alpha: float
how much prioritization is used
(0 - no prioritization, 1 - full prioritization)
"""
super(PrioritizedReplayBuffer,
self).__init__(size, frame_history_len, non_pixel_dimension,
add_non_pixel)
self.num_branches = num_branches
assert alpha > 0
self._alpha = alpha
it_capacity = 1
while it_capacity < size:
it_capacity *= 2
self._it_sum = SumSegmentTree(it_capacity)
self._it_min = MinSegmentTree(it_capacity)
self._max_priority = 1.0
def sample(self, batch_size, beta):
"""Sample a batch of experiences.
compared to ReplayBuffer.sample
it also returns importance weights and idxes
of sampled experiences.
Parameters
----------
batch_size: int
How many transitions to sample.
beta: float
To what degree to use importance weights
(0 - no corrections, 1 - full correction)
Returns
-------
obs_batch: np.array
Array of shape
(batch_size, img_c * frame_history_len, img_h, img_w)
and dtype np.uint8
act_batch: np.array
Array of shape (batch_size,) and dtype np.int32
rew_batch: np.array
Array of shape (batch_size,) and dtype np.float32
next_obs_batch: np.array
Array of shape
(batch_size, img_c * frame_history_len, img_h, img_w)
and dtype np.uint8
done_mask: np.array
Array of shape (batch_size,) and dtype np.float32
weights: np.array
Array of shape (batch_size,) and dtype np.float32
denoting importance weight of each sampled transition
idxes: np.array
Array of shape (batch_size,) and dtype np.int32
idexes in buffer of sampled experiences
"""
assert beta > 0
idxes = self._sample_proportional(batch_size)
weights = []
p_min = self._it_min.min() / self._it_sum.sum()
max_weight = (p_min * self.num_in_buffer)**(-beta)
for idx in idxes:
p_sample = self._it_sum[idx] / self._it_sum.sum()
weight = (p_sample * self.num_in_buffer)**(-beta)
weights.append(weight / max_weight)
weights = np.array(weights)
encoded_sample = self._encode_sample(idxes)
return tuple(list(encoded_sample) + [weights, idxes])
def store_frame(self, frame, non_pixel_feature):
"""Store a single frame in the buffer at the next available index, overwriting
old frames if necessary.
Parameters
----------
frame: np.array
Array of shape (img_h, img_w, img_c) and dtype np.uint8
the frame to be stored
Returns
-------
idx: int
Index at which the frame is stored. To be used for `store_effect` later.
"""
# if observation is an image...
if len(frame.shape) > 1:
frame = frame.transpose(2, 0, 1)
if self.obs is None:
self.obs = np.empty([self.size] + list(frame.shape),
dtype=np.uint8)
self.action = np.empty([self.size, self.num_branches],
dtype=np.int32)
self.reward = np.empty([self.size], dtype=np.float32)
self.done = np.empty([self.size], dtype=np.bool)
if self.add_non_pixel:
self.non_pixel_obs = np.empty(
[self.size, self.non_pixel_dimension], dtype=np.float32)
self.obs[self.next_idx] = frame
if self.add_non_pixel:
self.non_pixel_obs[self.next_idx] = non_pixel_feature
ret = self.next_idx
self.next_idx = (self.next_idx + 1) % self.size
self.num_in_buffer = min(self.size, self.num_in_buffer + 1)
return ret
def store_effect(self, idx, action, reward, done):
self.action[idx] = action
self.reward[idx] = reward
self.done[idx] = done
self._it_sum[idx] = self._max_priority**self._alpha
self._it_min[idx] = self._max_priority**self._alpha
def _sample_proportional(self, batch_size):
res = []
for _ in range(batch_size):
mass = random.random() * self._it_sum.sum(0,
self.num_in_buffer - 1)
idx = self._it_sum.find_prefixsum_idx(mass)
res.append(idx)
return res
def update_priorities(self, idxes, priorities):
"""Update priorities of sampled transitions.
sets priority of transition at index idxes[i] in buffer
to priorities[i].
Parameters
----------
idxes: [int]
List of idxes of sampled transitions
priorities: [float]
List of updated priorities corresponding to
transitions at the sampled idxes denoted by
variable `idxes`.
"""
assert len(idxes) == len(priorities)
for idx, priority in zip(idxes, priorities):
assert priority > 0
assert 0 <= idx < self.num_in_buffer
self._it_sum[idx] = priority**self._alpha
self._it_min[idx] = priority**self._alpha
self._max_priority = max(self._max_priority, priority)