|
8 | 8 |
|
9 | 9 | class Env: |
10 | 10 | """ |
11 | | - Represents an environment for multi env and multi agents. **Not gymnasium compatible**. |
| 11 | + Vectorized environments to run multiple independent environments simultaneously with multiple agents. |
12 | 12 |
|
13 | | - This class encapsulates the capability to reset, step, and close environments |
14 | | - within a simulation framework. It provides properties to access the runner, active |
15 | | - configs, simulation configuration, and other relevant components. The class is |
16 | | - designed to manage the lifecycle of simulation tasks. |
| 13 | + **NOT gymnasium compatible**. |
17 | 14 |
|
18 | 15 | Parameters: |
19 | 16 | config (Config): The config instance used for simulation |
20 | 17 | management. |
21 | | -
|
22 | | - Methods: |
23 | | - reset(env_ids: List[int] = None) -> Tuple[List, List]: Resets specified environments |
24 | | - and returns initial observations and task configs. |
25 | | - step(action: List[Union[Dict, OrderedDict]]) -> Tuple[List, List, List, List, List]: |
26 | | - Executes a single step in the environment using provided actions. |
27 | | - get_dt(): Retrieves the simulation timestep (dt). |
28 | | - get_observations() -> List | Any: Fetches observations from the simulation environment. |
29 | | - close(): Closes the simulation environment. |
30 | | - finished() -> bool: Checks if all tasks in the simulation are completed. |
31 | | -
|
32 | | - Properties: |
33 | | - runner: Provides access to the internal runner instance. |
34 | | - is_render: Indicates whether the environment is in a renderable state. |
35 | | - active_task_configs: Retrieves the currently active task configs. |
36 | | - simulation_app: Retrieves the simulation app instance. |
37 | 18 | """ |
38 | 19 |
|
39 | 20 | def __init__(self, config: Config) -> None: |
|
0 commit comments