Fix `vignettes/example-ratio-study.Rmd`

[wrridgeway]

Our AssessR vignette example-ratio-study.Rmd is how we explain our ratio stats. The Results starting here look scuffed.

What looks scuffed?

The actual values output for COD and PRB (and their CIs) are wildly unexpected and I'm not sure whether the root cause are scuffed methods, scuffed calculations, scuffed rendering, or some combination.

In the tabular result of Median ratios by sale price (which has columns for Decile and Sale Price), the 10% decile contains a median sale price of $1, suggesting we aren't doing sufficient trimming.

This vignette is using North Tri townships (New Trier, Palatine) and Tax Year 2020. Because 2020 was not a revaluation year for the North Tri, it's likely ratios would look "strange" for this year.

Suggested resolutions:

• change the tax_year, replacing 2020 with 2022.
• apply some trimming; for example, excluding sale prices <$10k.

Visual appearance of the "Median Sale Ratios: Open Data Sample" graph

This graph's x-axis is excellent and clear, but the y-axis is tough to read because it's in scientific notation.

Suggested resolution:

• enforce display range so that ratios are not in scientific notation.