Wrap Up


Congrats on working through all of the course materials. I hope you found this course to be useful and enjoyable. In this last section, I just wanted to highlight some options for continuing to learn R and work in this environment to analyze geospatial and earth science data.

Python and R

Currently, R and Python are the leading data science environments. You may find that some methods or tools are more developed in Python while others are more developed in R. I have found that I need to know how to work in both environments to complete my work; however, I am much more comfortable in R. Before diving into Python, you may want to see if an R wrapper has been created to use the Python modules or libraries you are interested in through R. This is generally accomplished using the reticulate package. For example, the keras R package allows you to use the Keras Python library via R. So, you can use this powerful deep learning tool set without learning Python. So, we would encourage you to explore options available in R that call to Python before tackling a new language. Note that this will require setting up Python on your machine along with Python virtual environments.

Dealing with Large Datasets

If you work with large data sets, you may want to explore some packages that are better at dealing with large tables than base R. Specifically, have a look at readr and/or data.table. If you work with processes that take a lot of computational time or resources, such as hyperparameter tuning, you may want to look into parallelization packages such as foreach, furrr, doParallel, and multidplyr.

Connecting to Other Geospatial Tools

There are several packages that allow you to connect to other geospatial software tools in R. Below are a few examples. For a summary of spatial data analysis packages in R, please visit this CRAN page.

  • RSAGA: connect to System for Automated Geoscientific Analysis (SAGA)
  • grass7: connect to Geographic Resources Analysis Support System (GRASS)
  • rgdal: connect to Geospatial Data Abstraction Library (GDAL)
  • rgee: connect to Google Earth Engine
  • RQGIS3: communication between R and QGIS 3.x.
  • R-ArcGIS Bridge: communication between R and ArcGIS Pro

Continued Learning

Fortunately, there are many free, online resources for learning R and many ways to troubleshoot errors and get help. The best way to learn R is to continue to practice. I hope you will consider integrating R into your workflows and continue building your skills in this powerful data science and computational environment.