Simple benchmark for RL algorithms.
python -m pip install -r requirements.txt
A logging farm game.
Run python main.py to start playing.
Control of game:
| Keyboard | Action |
|---|---|
| Up Arrow | Move cursor up |
| Down Arrow | Move cursor Down |
| Left Arrow | Move cursor left |
| Right Arrow | Move cursor right |
| c | Cut tree |
| n | Next year |
| 0 | Cut all trees with age 0 |
| 1 | Cut all trees with age 1 |
| 2 | Cut all trees with age 2 |
| 3 | Cut all trees with age 3 |
| 4 | Cut all trees with age 4 |
| 5 | Cut all trees with age 5 |
| 6 | Cut all trees with age 6 |
| 7 | Cut all trees with age 7 |
Final goal of game: Obtain as much profit(money) as possible in the limited number of years.
The simplest version. Trees with age 7 will seed seeds around them.
Environment is in Tree_env_1.py.
Training and Evaluation: python [algorithm].py
Added a threshold of CO2 absorption that needs to surpass.
Environment is in Tree_env_1.py.
Training and Evaluation: python [algorithm].py
Combined value of tree and CO2 into one reward.
Environment is in Tree_env_1.py.
Training and Evaluation: python [algorithm].py
Took soil fertility into consideration.
Environment is in Tree_env_1.py.
Training and Evaluation: python [algorithm].py
Integrated all ideas of previous versions.
Environment is in Tree_env_1.py.
Training and Evaluation: python [algorithm].py