求助：在Google Colab的R内核中安装GDAL及rgdal包

Hey there! I’ve helped many folks get spatial R packages like rgdal running in Colab, so let’s break this down simply for you and your students.

First, a quick key point: GDAL is a system-level library, not an R package, which means you can’t install it directly via R commands. Instead, you need to install it using Colab’s underlying Linux environment—and yes, Colab’s R kernel does support shell magic commands (similar to %%Shell in Python) using %%bash.

Here’s the step-by-step solution:

Step 1: Install System-Level GDAL Dependencies #

In a Colab R cell, run this bash magic command to install the required GDAL, Proj, and GEOS libraries (all critical for rgdal to compile and run):

%%bash
sudo apt-get update
sudo apt-get install -y libgdal-dev libproj-dev libgeos-dev

This updates the Linux package list and installs the core system libraries rgdal relies on.

Step 2: Install the rgdal R Package #

Once the system dependencies are set up, install rgdal as you normally would. In most cases, this straightforward command will work:

install.packages("rgdal", dependencies = TRUE, repos = 'http://cran.rstudio.com/')

If You Still Hit Errors: #

Occasionally, Colab might need explicit paths to the installed libraries. If the above command fails with a non-zero exit status, try specifying configure arguments to point Colab directly to the GDAL/Proj files:

install.packages("rgdal", 
                 dependencies = TRUE, 
                 repos = 'http://cran.rstudio.com/',
                 configure.args = c(
                   '--with-gdal-config=/usr/bin/gdal-config',
                   '--with-proj-include=/usr/include/proj',
                   '--with-proj-lib=/usr/lib/x86_64-linux-gnu'
                 ))

Verify the Installation #

After installing, load the package to confirm everything works:

library(rgdal)

If no errors appear, you’re all set—your students can now run their raster data processing scripts smoothly in Colab’s R kernel!

内容的提问来源于stack exchange，提问作者Syrph