Sandbox Upgrade 1
We want to create an environment to facilitate model tuning. Currently, we have two sandboxes 1_sf_modeling_sandbox.R, and 1_mf_modeling_sandbox.R.
We would like to create a single environment that can model single-family, multi-family, and condominium properties. The first iteration of this will be a Markdown file, with a planned transformation into a Shiny Application later.
The Markdown file should be able to accept parameter inputs, and output a 'modeling report' that contains the following outputs:
- Descriptive and summary statistics on the sample used in the modeling.
- A simplified regression table (see 'marginal effects' table).
- A table of sales ratio statistics.
- A map of outliers.
- Distribution of sales ratio.
Data construction will be tricky - ideally, this would be a standalone file that does not require other scripts to execute before functioning. However, that may not be efficient from a time perspective. For the first iteration, bring the data in using the quickest method possible, and we can then identify ways to improve flexibility from there.