VizX

Overview

databricks.labs.gbx.vizx renders rasters and turns Spark DataFrames into GeoDataFrames for interactive maps. It is tier-agnostic — the raster plotters take tile bytes (use them with pyrx or rasterx tiles), and the vector adapters take any Spark DataFrame.

These are driver-side, single-node helpers for inspecting results in a notebook — not distributed operations. They collect to the driver, so the vector adapters guard the collect with max_rows.

Opt-in extra

gbx.vizx requires geobrix[vizx], which pulls in matplotlib, geopandas, folium, and mapclassify. The package itself imports only matplotlib (raster rendering) and geopandas (the GeoDataFrame adapters); folium + mapclassify are used by plot_interactive / GeoDataFrame.explore() for interactive maps.

Key Features

• Raster plotting — plot_raster / plot_file render tile bytes or files (single-band, RGB, percentile stretch, coverage-depth, and mask-layer composites)
• VRT mosaic rendering — plot_mosaic renders a whole mini-COG VRT mosaic (from the cog_gbx vrtMosaic writer) as one memory-safe georeferenced image, with a bbox viewport and an optional h3 cell overlay
• DataFrame → GeoPandas adapters — as_gdf / cells_as_gdf / grid_as_gdf turn Spark geometry, H3-cell, and gridspec DataFrames into GeoDataFrames for .plot() / .explore()
• Multi-layer compositor — plot_static / plot_interactive overlay vector, raster, grid, and PMTiles layers in a single map
• PMTiles & COG viewers — plot_pmtiles, plot_cog, and pmtiles_info inspect and render archives inline
• Tier-agnostic, Python-only — works with pyrx or rasterx tiles; a driver-side notebook helper, not a SQL/Spark package

Installation

pip install "geobrix[vizx]"

From a Databricks notebook (commonly alongside the lightweight tier):

%pip install --quiet "geobrix[light_env6,vizx] @ file:///Volumes/<catalog>/<schema>/<volume>/geobrix-0.5.0-py3-none-any.whl"

Import

from databricks.labs.gbx.vizx import (
    plot_raster,
    plot_file,
    plot_mask_layers,
    plot_static,
    plot_interactive,
    plot_pmtiles,
    plot_cog,
    plot_mosaic,
    as_gdf,
    cells_as_gdf,
    grid_as_gdf,
)
from databricks.labs.gbx.pmtiles import pmtiles_info

What's in VizX

Sub-page	What it covers	When to use it
Multi-Layer Compositor	plot_static / plot_interactive with multiple layers; embed-size ladder; audit; simplify	Overlaying vector, raster, grid, and PMTiles layers in one map
PMTiles Viewers	pmtiles_info, plot_pmtiles, plot_cog	Inspecting and rendering PMTiles archives and COGs inline
Vector	plot_interactive (single-layer), plot_static (single-layer), as_gdf, cells_as_gdf, grid_as_gdf	Interactive and static single-layer maps; converting DataFrames to GeoDataFrames
Raster	plot_raster, plot_file, plot_mask_layers, plot_mosaic, escape hatches	Rendering raster tiles from bytes or disk, and whole VRT mosaics

Next steps

• Multi-Layer Compositor — layer types, embed-size ladder, ephemeral/durable simplify, scale guidance
• PMTiles Viewers — pmtiles_info, plot_pmtiles, plot_cog
• Vector — interactive and static single-layer maps, vector adapters
• Raster — raster rendering, escape hatches
• RasterX Function Reference
• STAC Client
• Helios notebooks — worked end-to-end example using plot_pmtiles, plot_cog, and pmtiles_info to render San Francisco tilesets inline.