Data Parsers

[miriamkw]

Enhance the data parsers:

• Add the Open APS parser
• Add the insulin column to the Ohio parser
• Implement the data cleaning steps and checks
• Implement (optional) derived features (this should be an input to the cli command):
  • In one script, we should first do the normal processing, save the df, then calculate derived features, and update the same saved file.
  • Hour, day, month, is_weekend ...
  • Insulin on board (use Jacobs et al. paper appendix for the formula)
  • Update README. Describe that certain features will take a long time to add.
• Implement so that several datasets can be used/merged for model training
  • Idea for solution: Configuration can have a list of datasets instead of one
  • PoC: Model, each id preprocessed individually, use transfer learning between the subjects (existing model?)

Note: This should be a helper for the cli instead of downstream in the parsers/preprocessors to avoid redundancy.