About this section

This section will cover some general concepts to be aware of when switching from ArcGIS to R. This is a (non exhaustive) list of the main differences we see between the two approaches:

• The fundamental building block of ArcGIS is a visual representation of your data (i.e. the map). The fundamental building block in R is your R-Script. The visual aspect makes ArcGIS more approachable, but also less scalable.
• When you acquire ArcGIS, you are served from a single source. In R, there are a plethora of developers creating packages for you to use, which can be quite overwhelming
• While there are always multiple ways to solve a problem in ArcGIS, there are usually many more ways to solve it in R. Finding the right way is not always easy.
• When you transition from a proprietary Software like ArcGIS to FOSS, you have to relearn many of the concepts because the proprietary software usually do not adhere to common standards.
• ArcGIS follows the concept of creating a tool for every situation you might encounter. With this approach, they provide a selection of > 1’000 geoprocessing tools to your disposal. Generally speaking, the authors of Packages will try to keep the number of functions as small as possible, since this reduces maintenance.

And just to get this out of the way: When is ArcGIS (or QGIS for that matter) preferable to using R? We will just cite Sébastien Rochette’s tweet

I would use QGIS for georeferencering and digitizing. - Georeferencing is when you get an aerial photo or an old paper map that you want to add in GIS - Digitizing is when you want to manually draw areas that you see on a digitized map Also, the reason I use to keep a QGIS installed, is to get the list of all projections and being able to search through the list to find the appropriate projection in case of problems with yours. Also, with {raster}, where you need to use the PROJ4 string, QGIS gives it to you.