Document the project data pipeline.

[nbanion]

In the Sphinx documentation, describe and diagram this project's pipeline from raw data sources to model predictions. This work will help new users contribute to the project, and it will help us organize project package code.

[nbanion]

@kelsonSS , I drafted a diagram of the inputs, models, and outputs based on some old slides. I'd like to leverage the team's knowledge to improve and fill out this diagram and also add some narrative.

Do you think some teammates could fill in the blanks or point to existing documentation?

Specifically, I'd like to:

• Annotate each arrow with the script or notebook that does the work.
• Identify arrows for processes that aren't implemented yet.
• Add boxes for interim datasets that we save to disk.
• Describe each of the inputs.
  • Source (organization and link).
  • High level description of subject matter.
  • Dimensions, years, geographic granularity / extent.
  • How we expect the data to change / grow.
  • Links to documentation for relevant code.
• Describe each of the models.
  • High level description of purpose.
  • Variables used.
  • Links to documentation for relevant code.

Doesn't have to be done all at once, but I think we can add details to this outline and develop a valuable resource over time.

[nbanion]

@kelsonSS , are we currently modeling with any data besides NFIRS, ACS, and Red Cross home visits?

Some other data sources that I've seen:

• American Housing Survey (AHS)
• American Red Cross Boundary Data
• Homeland Infrastructure Foundation Level Data (HIFLD)
• Social Vulnerability Index (SVI)