Module sac-pendulum.buffer
Classes
class Frame (state: Any, action: Any, reward: Any, state_next: Any, done: bool)
-
Frame(state, action, reward, state_next, done)
Expand source code
class Frame(NamedTuple): state: Any action: Any reward: Any state_next: Any done: bool
Ancestors
- builtins.tuple
Instance variables
var action : Any
-
Alias for field number 1
var done : bool
-
Alias for field number 4
var reward : Any
-
Alias for field number 2
var state : Any
-
Alias for field number 0
var state_next : Any
-
Alias for field number 3
class ReplayBuffer (capacity)
-
Expand source code
class ReplayBuffer: def __init__(self, capacity): self.capacity = capacity self.buffer: List[Frame] = [] def __len__(self): return len(self.buffer) def add(self, frame: Frame) -> None: self.buffer.append(frame) if len(self.buffer) > self.capacity: self.buffer.pop(0) def sample(self, sz) -> List[Frame]: return random.sample(self.buffer, sz)
Methods
def add(self, frame: sac-pendulum.buffer.Frame) ‑> None
def sample(self, sz) ‑> List[sac-pendulum.buffer.Frame]