The idea would be using https://github.com/ssenart/PyGazpar to send a weekly email with a recap of previous week consumption, a comparison with the weeks before and a diff with forecasted consumption. That later could be done with a simple time series model like that - http://www.martindaniel.co/roof/index.html
Questions:
- Should I store results in a DB and append incrementally, or re-run whole history each time?
- How to send the email? What format?
- How does GH action work?
To do:
The idea would be using https://github.com/ssenart/PyGazpar to send a weekly email with a recap of previous week consumption, a comparison with the weeks before and a diff with forecasted consumption. That later could be done with a simple time series model like that - http://www.martindaniel.co/roof/index.html
Questions:
To do: