CodingChallenge

AIMLAC Coding Challenge

Aims & Objectives

• Predict on site power generation at 30 minute intervals for at least the next 24 hours. The National Grid charges a penalty fee for any power generator who over estimates their power output and is unable to deliver.
• Automatically decide upon a price to sell any excess electricity to the National Grid. The price must be as profitable as possible while still being competitive with the general wholesale market. This must be specified at 30 minute intervals.
• Produce a regular report for management stating:
  • Revenue from selling electricity
  • Cost of imported electricity
  • Cost savings compared to buying from the National Grid at a rate of 15 pence per kilowatt hour
  • Carbon dioxide savings compared to buying from the National Grid. Note that the carbon intensity of the grid varies.

Installation

Currently, only building and installing from source is available.

Installing the anaconda environment

The conda environment is defined in environment.yml

• To create an environment use conda env create ./environment.yml
  • This will automatically install all the dependencies to the sjautobidder environment.
• Activate the environment with conda activate sjautobidder
• To delete an environment, use conda env remove -n envname

Building the frontend

For details about the frontend, see webpage/README.md

Installing the package

This would be used for deployment, but does not have a use while developing. Infact, installing the module might break testing with conflicted imported modules.

"conda activate sjautobidder;",
"python -m build .;",

# Optionally install with pip. --force-reinstall ensures it is installed
# even if the version has not been bumped.
"python -m pip install .\\dist\\sjautobidder-0.0.1-py3-none-any.whl --force-reinstall",

The distributable module is outputted to the dist/ folder, in both pip-installable .whl format and as a compressed .tar.gz folder.

Docker-compose

To start the docker container, make sure you have run the steps under Installing the package, then navigate to the root of the project and run:

docker build -t codingchallenge .
docker run -dp 80:80 codingchallenge

Linting

To lint the project:

• Activate the environment with conda activate sjautobidder
• Run the command prospector . -s high
• Also check comment style by running pydocstyle

Testing

To test the project:

• Activate the environment with conda activate sjautobidder
• Run the command pytest .

To generate a coverage report:

• For a terminal coverage report:

  • Use pytest . --cov-report term --cov=sjautobidder tests/

     ----------- coverage: platform win32, python 3.8.8-final-0 -----------
     Name                                               Stmts   Miss  Cover
     ----------------------------------------------------------------------
     sjautobidder\building_demand\__init__.py               0      0   100%
     sjautobidder\building_demand\energy_demand.py         31     13    58%
     sjautobidder\building_demand\energy_utils.py          24      0   100%
     sjautobidder\met_office_api\__init__.py                0      0   100%
     sjautobidder\met_office_api\api_interpolation.py      11     11     0%
     sjautobidder\met_office_api\api_utils.py             114    114     0%
     sjautobidder\solar_power\__init__.py                   0      0   100%
     sjautobidder\solar_power\solar_power.py               44     44     0%
     sjautobidder\solar_power\solar_utils.py               27     18    33%
     ----------------------------------------------------------------------
     TOTAL                                                251    200    20%

• To generate a cov.xml for IDE extensions:

  • Use pytest --cov-report xml:cov.xml --cov=sjautobidder tests/

Contributing

When committing code that may effect more than just your owned files, please open a pull request instead and tag the writer of the code you changed.

We use black as our formatting guide, and is recommended you run the formatter on your code before committing.

For doc comments, each function must be commented alongside a top-level module comment. The standard we have adopted is the Google Standard. Note that incorrect docs will be picked up as the linter and reported as errors. Please run the linter before committing!

Github actions will run pytest and linting on pull requests and commits, so pleasure be sure that your code conforms!

Pre-commit check list

• Use black formatter on your code.
• Run Pytest to ensure tests pass.
• Use the linter to ensure standards are followed, along with correct comments.
• If changing someone else's files, issue a pull request instead.