The R Programming Language

R is both a programming language and an environment. It is especially popular with data scientists who perform statistical analyses due to the large number of existing statistical packages available.

R-ArcGIS Bridge

Access to R from within ArcGIS Pro is provided via either the R-ArcGIS Bridge or the arcgis binndings package. These tools can be used to access and transform both local and remote data using the most common data types familiar to the R community. Data can be exchanged between ArcGIS Pro and the R environment and converted to and from R-specific data types such as spatial DataFrames. This allows the data from ArcGIS to be manipulated using R and written back to a geodatabase.

Big Data Connection

The big data connection allows ArcGIS Pro to read directly from collections of big data. A collection is a series of files. For example, a collection could consist of a large set of comma separated value files (CSV) or shapefiles stored in the cloud. The big data connection allows ArcGIS Pro to treat these large collections of files as a single file for the purpose of running an analysis.

Notebooks

A notebook is an interface for developing, running, and presenting the results of an analysis. Notebooks are often used to run Python code but can also be used for languages such as R.

A notebook present comments, code, and results as a series of cells. Individual cells can be run and rerun to fine-tune and adapt the analysis. Notebooks can be easily shared with others. A notebook can be run either in a standalone environment using the Jupyter Notebooks environment or hosted within ArcGIS Pro. Running a notebook within ArcGIS Pro automatically provides access to essential Esri libraries such as ArcPy. Other libraries such as Conda and machine learning packages can be installed in the ArcGIS Pro notebook environment.

Conclusion

A wide variety of tools exist for performing spatial data analyses such as ArcGIS geoprocessing tools, external statistical packages, and deep learning frameworks. Python is a key component that provides the thread to stitch all these tools together. R offers an alternative approach which may be useful for people who are more comfortable with that environment and the vast array of statistical packages it provides.