Code and data for "Reward design for deep Reinforcement Learning driven wind farm control: What matters for optimal performance"

Code and data accompanying the Energy and AI paper "Reward design for deep Reinforcement Learning driven wind farm control: What matters for optimal performance." Includes the analysis notebook and the NetCDF datasets that regenerate every figure in the paper, the SAC/PPO training and evaluation scripts used to produce the results, and a reference snapshot of the WindGym environment. Running reward_analysis.ipynb reproduces all figures from the bundled data; the training/evaluation scripts are provided for transparency and require the WindGym environment. See README.md for details.