rioxarray

When switching from rasterio to rioxarray, note that many file-specific attributes (like driver, tiled, compress, or interleave) are unused because rioxarray operates on xarray.DataArray abstractions which are not strictly tied to a single file on disk.

Profile Equivalences

Switching from rasterio to rioxarray is primarily done to integrate raster data into an xarray workflow.

Key advantages include:

• Unified Abstraction: xarray treats the raster dataset and the raster array as a single object, allowing you to use object attributes instead of passing separate arguments to functions.
• Advanced Computation: xarray provides built-in functions and supports backends like Dask for out-of-memory computation, cluster execution, or GPU acceleration.
• Dask Support: rioxarray supports basic dask features during I/O operations.

Important Considerations:

• Information Loss: Be aware of potential information loss when working with xarray abstractions.
• Naming Confusion: A rasterio Dataset and an xarray Dataset are distinct objects; be careful when using these overlapping terms in your code.

You can install the stable release of rioxarray from PyPI using pip.

pip install rioxarray

Before submitting a pull request, ensure the following requirements are met:

1. Include Tests: Every PR must include corresponding tests.
2. Update Documentation: If adding functionality, update the docs. This includes adding a docstring to the new function and adding the feature to the list in README.rst.
3. Python Compatibility: The code must work for Python versions 3.12 through 3.14.

To use rioxarray features on xarray objects, you must import the rioxarray package. This registers the .rio accessor, which provides geospatial capabilities like clipping, merging, and reprojecting rasters.

import rioxarray

You can install rioxarray directly from the GitHub repository using pip.

python -m pip install git+git://github.com/corteva/rioxarray.git#egg=rioxarray

To install rioxarray using conda, it is recommended to use the conda-forge channel. It is best practice to install the package into a new environment rather than your base environment to ensure stability and easier debugging.

Warning: Avoid using pip install inside a conda environment if possible. If a package is missing from conda-forge, consider submitting a recipe to the conda-forge community instead.

conda config --prepend channels conda-forge
conda config --set channel_priority strict
conda create -n rioxarray_env rioxarray
conda activate rioxarray_env

When loading netCDF files using standard xarray methods, it is recommended to use decode_coords="all". This ensures that grid mapping variables are loaded into coordinates, which is necessary for compatibility with rioxarray operations.

import xarray

xds = xarray.open_dataset("file.nc", decode_coords="all")

For specific, common troubleshooting and implementation questions, rioxarray users often refer to these documented StackExchange solutions:

• Extracting data within geometry (shape)
• Converting NetCDF dataset array to GeoTiff
• Adding projection to a NetCDF file (Satellite)
• Creating a new raster TIFF file masked based on a GeoJSON file
• Masking NetCDF time series data from a shapefile
• Extracting data from a raster at a specific point
• Converting a raster to CSV with lat, lon, and value columns

To extend xarray objects with rioxarray functionality, you must import rioxarray. This activates the .rio accessor on xarray.DataArray and xarray.Dataset objects, allowing you to access geospatial operations like coordinate management, clipping, and reprojecting.

import rioxarray
# Now xarray objects have the .rio accessor available

To set up a local development environment, clone the repository and install it in editable mode with the dev group.

git clone git@github.com:corteva/rioxarray.git
cd rioxarray
python -m pip install -e . --group dev