Separate functionality of get_df() into a CLI script

[jvivian]

After talking with @AaronCooke2718 I would like to implement #56 next when I have the time. The goal is to allow someone to run the model with any time-series input, not specific to our project.

Doing so will require replacing get_df() with two input files:

1. time series x variables matrix (after processing / before normalization -- currently what get_df() generates)
2. A variables x metadata matrix which includes a factor column that describes the variable-factor relationship

get_df() is referenced in a lot of places, so will need testing of all streamlit pages prior to merge

[jvivian]

Implemented in #74 viaDataLoader class @AaronCooke2718