This repository contains the example implementation for how the optimization challenge code will be executed at the end of the competition. Competitors are required to only implement a single method propose_state within the file battery_controller.py.

This code exists to make it easy for competitors to test their solutions.

• Docker
• Python 3.6 (for local execution instead of on Docker)

1. Clone this repository
2. Add the data from the competition to the data folder. The script expects a folder named submit and a file called metadata.csv in the data directory.
3. Copy your implementation of battery_controller.py into simulate/battery_controller.py
4. Execute the run script: ./run.sh
5. Your results will be stored and timestamped in the all_results folder each time you execute run.sh (Note: results.csv in output will be overwritten on each subsequent run.)

Note for Windows Users: We will accept a pull request with a run.bat script that works on Windows machines.

The only supported execution is within Docker. However, if you want to run the Python code locally rather than in a Docker container, you can still use the entrypoint.sh script on unix based system. You may need to install coreutils on your system in order to use the timeout command. For OSX you can run brew install coreutils and then update entrypoint.sh to call gtimeout instead of timeout. Windows users will need to create their own script, although for basic purposes just running python simulate/simulate.py should be sufficient.

The public leaderboard just presents the mean of your results so far. Simply submit the most recent results from the all_results folder.

You are required to submit your code in order to be considered for a prize. To do so, you must create a zip archive containing ONLY the assets folder and the battery_controller.py file. These are the only components you may submit. Other folders and files will be ignored. Instructions for submitting are on the competition page.