GraphEnvs
Graph Reinforcement Learning (RL) Environments
Installation
First install torch and pytorch geometric. The simply use pip to install graph_envs:
pip install graph-envs
Supported Environments
GraphEnvs-Basic:
| Environment | Developed | Action Space |
|---|---|---|
| Shortest Path | ✅ | $v \in \mathcal{V}$ |
| Steiner Tree | ✅ | $e \in \mathcal{E}$ |
| MST | ✅ | $e \in \mathcal{E}$ |
| Minimum Vertex Cover | ✅ | $v \in \mathcal{V}$ |
| TSP | ✅ | $v \in \mathcal{V}$ |
| Longest Path | ✅ | $v \in \mathcal{V}$ |
| Largest Clique | ✅ (Min Vertex Cover) | $v \in \mathcal{V}$ |
| Densest Subgraph | ✅ | $v \in \mathcal{V}$ |
| Node Coloring | 🛠️ | $(v, c) \in \mathcal{V} \times \mathbb{Z} $ |
GraphEnvs-Extended:
| Environment | Developed | Action Space |
|---|---|---|
| MultiCast Routing | ✅ | $e \in \mathcal{E}$ |
| Distribution Center Selection | ✅ | $v \in \mathcal{V}$ |
| Persihable Product Delivery | ✅ | $v \in \mathcal{V}$ |
| Public Transport Navigation | 🛠️ | - |
Example
import gymnasium as gym
import graph_envs
import numpy as np
env = gym.make('LongestPath-v0',
n_nodes=10,
n_edges=20,
weighted=True,
is_eval_env=True,
parenting=2
)
for sd in range(0, 10):
print(f'===== {sd} =====')
obs, info = env.reset(seed=sd)
mask = info['mask']
done = False
while not done:
valid_actions = mask.nonzero()[0]
action = np.random.choice(valid_actions)
obs, reward, done, _, info = env.step(action)
print('Valid actions:', valid_actions, ' Action:', action, ' Reward:', reward, ' Done:', done)
mask = info['mask']
print(info['solution_cost'], info['solved'], info['heuristic_solution'])
Environment Details:
- Shortest Path: The goal is to find the shortest path from the source node to the target node. At each step, an edge is added to the path. The episode is over when we reach the target node.
- Steiner Tree: The goal is to find the tree with the minimum weight that connects a source node to a number of destination nodes.
1 maintainer
Release files for graph-envs 0.0.55
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| graph-envs-0.0.55.tar.gz | 15.8 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| graph_envs-0.0.55-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 41.9 kB
Release files / graph-envs-0.0.55.tar.gz
| Download URL | graph-envs-0.0.55.tar.gz |
|---|---|
| Size | 15.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
ba1e96525a132d5933fecd11981b85255125134f14d3d7980cb98eaccb7cdb82
|
|
BLAKE2b-256 checksum How to use checksums |
7d7da648ac194d1e5bb28ba15ec5caf943481f9b3265f4b5e18c6e6295abfb8b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/4.0.2 CPython/3.10.11
|
Release files / graph_envs-0.0.55-py3-none-any.whl
| Download URL | graph_envs-0.0.55-py3-none-any.whl |
|---|---|
| Size | 26.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
6f1a0d9af066301aff7e0de0ed541f27c15fae183882cf4e09ab5ce37fa8bd70
|
|
BLAKE2b-256 checksum How to use checksums |
2ac02de4e3f7c3a89e488d7843113a3f1840ace448d65aa3957b67b3d5f0304a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/4.0.2 CPython/3.10.11
|
