WherobotsDB Python API

Wherobots notebooks and job runs come with the WherobotsDB Python packages installed, so you don’t need to install anything. Most notebooks import sedona.spark and create a Sedona context first:

from sedona.spark import *

config = SedonaContext.builder().getOrCreate()
sedona = SedonaContext.create(config)

from sedona.spark import * brings in SedonaContext, every DataFrame spatial function such as ST_Contains, the spatial statistics functions, and the geometry and raster column types. You can also import each name from the module shown on its reference page.

Packages and modules

Module	Reference	Contents
`sedona.spark`	SedonaContext	Create a Spark session with the spatial functions and types registered.
`sedona.spark.sql.st_constructors`	DataFrame constructors	Build geometry columns from coordinates, text, and binary formats, such as `ST_Point` and `ST_GeomFromWKT`.
`sedona.spark.sql.st_functions`	DataFrame functions	Measure, edit, transform, and index geometry columns, such as `ST_Buffer` and `ST_Area`.
`sedona.spark.sql.st_predicates`	DataFrame predicates	Test the spatial relationship between two columns, such as `ST_Contains` and `ST_Intersects`.
`sedona.spark.sql.st_aggregates`	DataFrame aggregates	Aggregate geometry columns, such as `ST_Union_Agg` and `ST_Envelope_Agg`.
`sedona.spark.stats`	Spatial statistics	DBSCAN clustering, local outlier factor, Getis-Ord hotspot detection, distance-band weights, spatially stratified sampling, and Moran’s I.
`sedona.spark.geopandas`	GeoPandas API	`GeoDataFrame`, `GeoSeries`, and file readers that follow the GeoPandas API and run on Spark.
`sedona.spark.stac`	STAC client	Search STAC APIs and load items into Spark DataFrames.
`wherobots.vtiles`	Vector tiles	Generate vector tiles and PMTiles archives from Spark DataFrames.
`wherobots.matcher`	Map matching	Match GPS points to road network edges.
`wherobots_gl`	Map widget	Render the Wherobots map inline in a notebook.

Each DataFrame function page links to the SQL function of the same name in the WherobotsDB SQL reference, which describes its behavior in more detail. Scala and Java code can call the spatial statistics, map matching, and vector tiles functions on these pages under camelCase names, for example LocalOutlierFactor.localOutlierFactor for local_outlier_factor.

Older import paths without .spark, such as sedona.sql.st_functions or sedona.stats.clustering.dbscan, still work but are deprecated. Use the sedona.spark paths on these pages.

​Packages and modules

Packages and modules