Create single location end-to-end forecasting CLI and associated container image

[dylanhmorris]

This PR:

• Creates a script (forecast_state.py) that runs a forecast end to end (from data prep through post-processing) and writes the results (including intermediate products) to designated locations.
• Creates an OCI container to allow that script to be run easily on cloud machines (e.g. Azure batch nodes).
• Adds scripts under nssp_demo/batch to set up pools and jobs.
• Transitions from using renv to manage R deps to creating a skeleton package (hewr) with all needed R deps as Imports, and installs it in as part of the container build.
• Does some cleanup on scripts
• Fixes terminology around "ed visits" for clarity / precision.
• Reduces default infection feedback priors.
• All of the above has been tested in Azure batch.

[damonbayer]

I left many comments, but no mandatory changes. Let me know if you would like me to address any of them myself.

[dylanhmorris]

Merging this @damonbayer. Thanks for the review / comments.