Future of get_eurostat_geospatial

[pitkant]

Executive summary

Two alternative paths for package development:

• Retire all internal mapping related functions (get_eurostat_geospatial()) and provide improved documentation / vignette and potentially some helper functions for giscoR package.
  • Would make eurostat package more lightweight, more focused on just retrieving statistical data
  • Would make it maybe less useful for less experienced end users?
• Retire only code, examples and tests related to using sp package
  • Provides backward compatibility, gives more time to soft deprecate functions if we so wish

Deadline

October 2023

Background

To make map visualization easy for end users eurostat has had built-in functions to visualise Eurostat data on a map and has included sf and sp packages as Imports. However, in 2023 rgdal, rgeos and maptools packages will be retired and sp package will evolve, having sf at minimum as a weak dependency.

More information from this message and in the link:

The legacy packages maptools, rgdal, and rgeos, underpinning this package
will retire shortly. Please refer to R-spatial evolution reports on
https://r-spatial.org/r/2023/05/15/evolution4.html for details.

Other considerations

• Directly affects issue / potentially solves Specify that libgdal-dev is necessary to install on Debian/Ubuntu or make the sf package optional #213
• Partially related to / potentially solves issues Problem accessing geospatial data in R "no access to ec.europe.eu." #222, Error installing package sf. #233, Duplicated variable columns in geospatial data #240
• Probably requires some edits to eurostat cheatsheet

[pitkant]

Addressed in v4-dev branch and PR #264. Relying on giscoR package in get_eurostat_geospatial functionalities will be the way to go in the future.

I will close this issue when v4 is released

[pitkant]

Closed with the CRAN release of package version 4.0.0