Sinergym Tutorial

This tutorial will walk you through the steps to training a RL agent using the HNP package inside the Sinergym environment.

Tip

Make sure you follow the installation instructions in Sinergym repository or documentation to set up your Sinergym environment correctly.

1. Navigate to /examples/sinergym, and you will see the following files:

   • sinergym_config.yaml: Agent and environment configuration file for HNP.

   • train_sinergym.py : The RL agent training script.

2. Run the following command to start training:

   python train_sinergym.py sinergym_config.yaml
   

3. Once the training completes, the rewards will be saved to training_results/yyyy_mon_dd/results_H_M_S/rewards.npy. You can use examples/plot_results.py to plot the rewards by running:

   python plot_results path/to/rewards.npy
   

   Alternatively, if you want to compare the rewards between different runs/agents, you can append the second rewards path to the above command to plot the difference.