Atlas Brasil Explorer

This is a repository for the Atlas Brasil Shiny Dashboard. You can access the interactive app here.

Repository Structure

This repository is organized and structured similarly to an R package, although it is not a package itself. I do so to keep the workflow efficient. The raw, uncleaned data from Atlas Brasil, downloaded from their site, is inside the /data-raw folder. This folder also contains the R scripts used to clean the data. The original files and dictionaries are compressed in zip format. Additionally, I created the Excel sheets named dictionary_* since I had to translate all variable names and descriptions.

The cleaned data is exported to the /data folder. Tables are exported to csv and the sf objects (shapefiles with data) are exported to qs for both size and efficiency reasons¹.

The actual scripts used inside the app reside in the /R folder. The interactive map is made with tmap and the dumbbell plot is made with plotly. Data manipulation inside the app is minimal. The overall user interface (UI) and JavaScript code for the map was heavily inspired by the SuperZip example from the Shiny Gallery.

To ensure future compatibility, I utilize the renv package, which helps maintain the functionality of the app even if the packages used undergo changes.

Footnotes

1. I did experiment with gpkg and rds with different types of compression as well and found the qs extension to work better. For more information on reading and writting qs files check the qs repository. ↩