Skip to main content
Databricks Labs

GeoBrix

High-performance spatial processing library for Databricks

RasterX

Satellite imagery, elevation models, and gridded data — reprojection, terrain analysis, spectral indices, XYZ/PMTiles tiling, and H3/quadbin aggregation.

Learn more →

GridX

Discrete global grid indexing — British National Grid, CARTO quadbin, and custom user-defined grids: cell math, tessellation, and grid-aware aggregation.

Learn more →

VectorX

Mapbox Vector Tile encoding, TIN elevation surfaces, and legacy Mosaic geometry migration to Databricks spatial types.

Learn more →

Powerful Readers & Writers

Automatically registered Spark readers and writers for Shapefile, GeoJSON, GeoPackage, GeoTIFF, PMTiles, and more.

Learn more →

Multi-Language APIs

Native Scala implementation with Python and SQL bindings for maximum flexibility.

Learn more →

Databricks Native

Built exclusively for Databricks Runtime, integrated with product spatial functions.

Learn more →

Start lightweight — the recommended raster tier

The lightweight tier (pyrx) runs the full GeoBrix raster API on pure Python + rasterio: a single wheel, no JAR and no init script, and it works everywhere — serverless, standard/shared clusters, ARM, and Lakeflow declarative pipelines. VectorX is likewise available lightweight (pyvx), as is all of GridX — BNG, quadbin, and custom grids (pygx). The heavyweight Scala/GDAL tier is there when you need full GDAL/OGR.

Compare Execution Tiers →

Quick Start

Get up and running with GeoBrix in minutes:

# Install the lightweight wheel (single library, no JAR, no GDAL)
# Stage the wheel on a Unity Catalog Volume, then install the quoted PEP 508 named form:
%pip install --quiet "geobrix[light] @ file:///Volumes/<catalog>/<schema>/<volume>/geobrix-<version>-py3-none-any.whl"

# Import and register functions
from databricks.labs.gbx.pyrx import functions as rx
rx.register(spark)

# Read and process geospatial data
rasters = (spark.read.format("binaryFile").load("/data/rasters")
           .select(rx.rst_fromcontent("content").alias("tile")))
metadata = rasters.select(
    rx.rst_boundingbox("tile").alias("bbox"),
    rx.rst_metadata("tile").alias("metadata")
)

Ready to get started?

Install GeoBrix on your Databricks cluster and unlock powerful geospatial capabilities.

Installation Guide