---
id: odc-geo
version: "0.5.3"
license: unclear
license_treatment: permissive
maintenance: active
---

# odc-geo — Geometry Classes and Operations (opendatacube)

License: permissive · Maintenance: active · Downloads: 168.7K/mo

## What it is and what it does

odc-geo is a geospatial library that wraps shapely geometries with CRS awareness from pyproj, preventing accidental operations between geometries in different projections. It provides GeoBox, an abstraction for geo-registered pixel planes with linear world-coordinate mappings, and integrates with xarray to attach and preserve geospatial metadata through data transformations.

The library was extracted from Open Datacube to serve projects working with geo-registered raster data, satellite imagery, and other geospatial datasets. It enables robust handling of projection information alongside geometry operations, making it easier to work with multi-CRS workflows and to attach geospatial context to xarray Datasets and DataArrays in ways that survive serialization and common data operations.

Use it for:

- Prevent projection mismatches by refusing geometry operations across different CRS and providing explicit reprojection.
- Define and work with geo-registered pixel grids (GeoBox) for satellite or raster imagery analysis.
- Attach and preserve geospatial metadata on xarray objects through mathematical operations and format conversions.
- Build geospatial data pipelines where projection consistency is enforced at the library level.
- Work with geospatial information in ways that survive serialization to zarr, netcdf, and GeoTIFF formats.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides projection-aware geometry classes combining shapely shapes with pyproj CRS support, plus GeoBox for geo-registered raster data and xarray integration for geospatial metadata.

Yes. Active maintenance, no known vulnerabilities, low install friction, and a permissive license make it safe to adopt. Install if you work with geo-registered raster data, need CRS-aware geometry operations, or want to attach robust geospatial metadata to xarray objects. The library is mature and widely used in geospatial data science.

## Install

pip install odc-geo
uv add odc-geo
poetry add odc-geo

## Installing odc-geo

Before you install: Low friction: pure Python wheel with six well-established dependencies (affine, cachetools, numpy, pyproj, shapely, xarray). Active maintenance with a release 29 days ago and recent commits.

License in practice: Permissive license allows commercial and private use without restriction.

Quickstart:

pip install odc-geo

from odc.geo.geom import Geometry
from pyproj import CRS

geom = Geometry(shape, crs=CRS.from_epsg(code))
geom_reprojected = geom.to_crs(CRS.from_epsg(target_code))

Requires Python 3.10 or later.

Verify before relying:

- Whether GeoBox's linear mapping supports all common geospatial coordinate systems or has known limitations.
- Performance characteristics when working with large raster datasets or many geometries.
- Specific EPSG codes or coordinate systems supported by the GeoBox implementation.

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 168.7K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

projection-aware geometry, geospatial raster metadata, geobox pixel coordinate mapping, xarray geospatial accessor, crs-aware shapely geometries, geo-registered data handling, opendatacube geometry utilities, geospatial, raster-data, crs-projection

[View on SkillFed](https://skillfed.io/packages/odc-geo) · [View on PyPI](https://pypi.org/project/odc-geo/)