snowdj/vse-risk-tool

VSE COVID Risk Tool Visualization Code

vse-risk-tool

Code repository for the VSE COVID Risk Tool, hosted online at https://baylislab.shinyapps.io/risk-tool-v2/.

Purpose

VSE COVID-19 Risk Tool can be used to compare some of the risks and benefits of re-opening different sectors of the BC economy. This repository contains the code necessary to produce a Shiny app with two main visualizations:

1. An interactive scatterplot comparing measures of economic productivity with a risk index for occupations in BC sectors.
2. A flexible table to show measures of risk, economic productivity, and a range of other relevant measures by disaggregated occupations and subsectors. This table can respond to selections made in the scatterplot.

Usage

1. Open vse-risk-tool.Rproj in RStudio.
2. Execute Run App in RStudio.

File structure

• data/input/ contains tidy data aggregated to its specific level for each province, for example:
  • /BC/ind_3_digit.csv: Measures of centrality for each subsector
  • /BC/ind_2_digit.csv: Share of GDP and employment for each sector
  • /BC/occ_4_digit_job.csv: Occupation risk levels for each subsector
  • /BC/occ_4_digit_hh.csv: Risk levels associated with the household for each subsector
  • /BC/occ_2_digit_desc.csv: Two digit occupation code (NOC47) and its description
  • /BC/risk_index_coefficients.csv: Risk coefficients used in VSE Risk Tool
• data/output/ contains the input data merged for each province to be used in the VSE Risk Tool

Note on build-data.R

By default, this tool produces these figures using BC-specific data in data/, which are produced by build-data.R. build-data.R builds the BC-specific on source data on occupation risk and economic value from outside of the repository. Users with access to the BC-specific data can run build-data.R by adding their location for the cleaned BC-specific data to the RAW_POSSIBLE character vector in setup.R.

Most users, however, should consider this repository to be a example of a useful way to visualize the underlying data. However, repurposing this code for other regions will require significant additional data work, as it is only currently designed to handle BC-specific data.