- A Gmail account with archived mails from the Probo mailing system
- An app password to login to Gmail.
If you don't have above, you can simply play around by selecting to use cached data when prompted.
> python main.py produces the current_predictions.png - and overview of the probability of an event happening the next 14 days after the last event.
- Parsing receipts for laundry bookings from the digital owner's association management tool Probo.
- Parsing my Gmail using
imaplibfor email confirmations for laundry slots. Only mails NOT deleted (not archived) will be retrieved. - Currently classifies a laundry-session as a binary event. (Future work would be playing around with the length of the sessions.)
- Fitting a logistic regression model on the data to obtain the probabilities for a laundry session to take place
$k$ days after the last session.
There's a lot of competition for the laundry machines in my apartment building. I wanted to know when the best time to do laundry was.
- Find methods to interpolate gap in data from my semester abroad (January 22' to May 22').
- Lookup calendar to find best slot for laundry session.
- Play around with time series models
- Make script update cache with newer data than latest entry