swarmrl.value_functions.expected_returns Module API Reference¶
Module for the expected returns value function.
ExpectedReturns
¶
Class for the expected returns.
Source code in swarmrl/value_functions/expected_returns.py
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 |
|
__call__(rewards)
¶
Call function for the expected returns. Parameters
rewards : np.ndarray (n_time_steps, n_particles, dimension) A numpy array of rewards to use in the calculation.
Returns¶
expected_returns : np.ndarray (n_time_steps, n_particles) Expected returns for the rewards.
Source code in swarmrl/value_functions/expected_returns.py
41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 |
|
__init__(gamma=0.99, standardize=True)
¶
Constructor for the Expected returns class
Parameters¶
gamma : float A decay factor for the values of the task each time step. standardize : bool If True, standardize the results of the calculation.
Notes¶
See https://www.tensorflow.org/tutorials/reinforcement_learning/actor_critic for more information.
Source code in swarmrl/value_functions/expected_returns.py
19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 |
|