Getting Started With RGEE

I need to get some data hosted through Google Earth Engine. I’ve played around with the javacript and python libraries, but at heart I’m an R developer and would prefer to use earth engine in R if I can.

This post is going to assume you have some python experience and can set up a python virtual environment of some sort. This is also done on a MacOS in zsh. Getting a python environment setup is a little different on Windows.

Note: I use VScode and call R using Radian a python package. Because of this my default python environment was the environment that radian is in. I basically could not change that environment no matter what I tried. I would also note that it seems like many folks have trouble setting their environment within reticulate.

Setup a python environment and install rgee

To get rgee working we need to get a python environment set up. To start from the terminal in bash or zsh (not an R terminal) make sure python is installed:

❯ python --version
Python 3.12.1

I have python 3.12.1 installed, the latest versions as of the writing of this post. Then you can create a venv from whatever directory you wish (cd there first) and initialize it.

python -m venv rgee_env
source rgee_env/bin/activate

You should have (rgee_env) in your terminal prompt if the environment was successfully created. Now we need to install the dependencies earthengine-api and Numpy. From the terminal again (not the R terminal):

pip install earthengine-api
pip install numpy

You should have all the correct python libraries to run rgee from R.

Setup Google Earth Engine Project

Create or use and existing account at the Google Earth Engine website and create a project. Make note of the project ID.

library(rgee)
library(reticulate)
library(tidyverse)
library(terra)

Login to Earth Engine

From here I am not going to evaluate any of these blocks because you cannot ee_Initialize() or get images that are stored together.

Here we initialize my google cloud account and then make sure my credentials are good.

ee_Initialize(user = 'mschmidty@gmail.com', drive = TRUE)
ee_check_credentials()

Check Earth Engine and Python packages

Then we check that Earth Engine python packages are installed.

ee_check()
ee_check_python_packages()
ee_check_packages()

Make a Map

Next we tell Google Earth Engine what collection we would like to use, set a geometry and projection. The geometry and projection are more so that I can download this image.

srtm <- ee$Image("USGS/SRTMGL1_003")
geometry <- ee$Geometry$Rectangle(
  coords = c(-108.274715, 37.076888, -107.213842, 37.965651),
  proj = "EPSG:4326",
  geodesic = FALSE
)

Download the image as a {terra} SpatRast

This step downloads the above specified image to a folder on google drive. ee_monitoring(task=as_rast) tells us the status of the download while the next step rast_col_rd<-columbine_rd|>ee_utils_future_value() adds the image to memory. The easiest way to use the image from here is to save the raster to memory and then call it from there.

columbine_rd <- ee_as_rast(
  image = srtm,
  via = "drive",
  region = geometry,
  lazy = TRUE
)
ee_monitoring(task=as_rast)
rast_col_rd<-columbine_rd|>ee_utils_future_value()