title: SpatialBench Datasets and Generators

SpatialBench Datasets and Generators

This page describes the SpatialBench datasets and shows you how to use the generators to create the spatial tables.

SpatialBench is a geospatial benchmark designed for evaluating and optimizing spatial query performance in data systems. Inspired by the Star Schema Benchmark (SSB) and the New York City Taxi and Limousine Commission (NYC TLC) dataset, SpatialBench blends realistic urban mobility scenarios with standardized benchmarking practices.

The benchmark adopts the familiar star schema structure from SSB, augmented with spatial attributes such as pickup and dropoff points, spatial polygon boundaries for zones, and building footprints. These spatial enhancements allow SpatialBench to effectively test geospatial operations, including spatial joins, distance-based queries, spatial aggregations, and point-in-polygon analyses.

By combining the systematic approach of SSB with authentic, real-world scenarios drawn from NYC TLC data, SpatialBench provides meaningful and practical benchmarks relevant to urban mobility and spatial analytics workloads.

Data model

SpatialBench tables:

Trip (Fact Table): Records individual trips, including spatial attributes (pickup and dropoff points), trip fare, distance, duration, and timestamps for pickup and dropoff.
Customer: Represents customers who book trips.
Driver: Represents drivers who fulfill trips.
Vehicle: Details about vehicles used for trips.
Zone: Polygon boundaries representing city areas or zones.
Building: Polygon footprints representing building locations, types, and names.

Table	Type	Abbr.	Primary Role	Spatial Attributes	Size per Scale Factor (SF)
Building	Dimension	b_	Polygon footprints representing building locations	Polygon footprints	20K × (1 + log₂(SF))
Customer	Dimension	c_	Represents customers	None	30K × SF
Driver	Dimension	s_	Represents drivers	None	500 x SF
Trip	Fact Table	t_	Records individual trips	Pickup/Dropoff Points (location)	6M × SF
Vehicle	Dimension	v_	Details about vehicles	None	100 x SF
Zone	Dimension	z_	Polygon boundaries for city zones	Polygon boundaries	Tiered by SF range (see below)

Zone Table Scaling

Scale Factor (SF)	Zone Subtypes Included	Zone Cardinality
[0, 10)	microhood, macrohood, county	156,095
[10, 100)	+ neighborhood	455,711
[100, 1000)	+ localadmin, locality, region, dependency	1,035,371
[1000+)	+ country	1,035,749

schema

Geographic Coverage

Spatial Bench's data generator uses continent-bounded affines. Each continent is defined by a bounding polygon, ensuring generation mostly covers land areas and introducing the natural skew of real geographies.

Bounding polygons:

Region	Bounding Polygon
Africa	`POLYGON ((-20.062752 -40.044425, 64.131567 -40.044425, 64.131567 37.579421, -20.062752 37.579421, -20.062752 -40.044425))`
Europe	`POLYGON ((-11.964479 37.926872, 64.144374 37.926872, 64.144374 71.82884, -11.964479 71.82884, -11.964479 37.926872))`
South Asia	`POLYGON ((64.58354 -9.709049, 145.526096 -9.709049, 145.526096 51.672557, 64.58354 51.672557, 64.58354 -9.709049))`
North Asia	`POLYGON ((64.495655 51.944267, 178.834704 51.944267, 178.834704 77.897255, 64.495655 77.897255, 64.495655 51.944267))`
Oceania	`POLYGON ((112.481901 -48.980212, 180.768942 -48.980212, 180.768942 -10.228433, 112.481901 -10.228433, 112.481901 -48.980212))`
South America	`POLYGON ((-83.833822 -56.170016, -33.904338 -56.170016, -33.904338 12.211188, -83.833822 12.211188, -83.833822 -56.170016))`
South North America	`POLYGON ((-124.890724 12.382931, -69.511192 12.382931, -69.511192 42.55308, -124.890724 42.55308, -124.890724 12.382931))`
North North America	`POLYGON ((-166.478008 42.681087, -52.053245 42.681087, -52.053245 72.659041, -166.478008 72.659041, -166.478008 42.681087))`

continents

Distribution Options

By default, SpatialBench generates points using continent-bounded affines with a Hierarchical Thomas distribution for the trip and building tables.

For more realism, you can choose from a variety of spatial distributions when generating tables:

Uniform: Evenly spread points in the unit square.
Normal: Gaussian spread around a mean with configurable variance.
Diagonal: Points concentrated along the y=x diagonal with configurable buffer.
Bit: Recursive grid-like pattern controlled by probability and bit depth.
Sierpinski: Self-similar fractal pattern for highly skewed coverage.
Thomas: Clustered distribution with realistic hotspots and heavy-tailed skew.
Hierarchical Thomas: Multi-level clustering (cities → neighborhoods → points), useful for mimicking urban settlement patterns.

These options let you tailor the spatial skew to your benchmarking needs.

See the SpatialBench Data Distributions page to learn more about the supported spatial distributions, the parameters that control them, and how they impact the data.

Data generators

You can generate the tables for Scale Factor 1 with the following command:

spatialbench-cli -s 1 --format=parquet --output-dir sf1-parquet

You can also generate data directly to Amazon S3 by providing an S3 URI:

spatialbench-cli -s 1 --format=parquet --output-dir s3://my-bucket/sf1-parquet

See the Quickstart for details on configuring AWS credentials.

Here are the contents of the sf1-parquet directory:

building.parquet
customer.parquet
driver.parquet
trip.parquet
vehicle.parquet
zone.parquet

See the README for a full description of how to use the SpatialBench data generators.

Data sizes

Here are the uncompressed Parquet file sizes of the tables for some different scale factors:

Category	SF1	SF10	SF100	SF1000
Zone	1.3 GB	2.0 GB	5.4 GB	5.7 GB
Trip	471.1 MB	5.0 GB	50.4 GB	512.7 GB
Building	2.4 MB	10.2 MB	18.0 MB	0.03 GB
Customer	2.5 MB	23.1 MB	227.1 MB	2.2 GB
Driver	0.04 MB	0.4 MB	4.0 MB	0.03 GB
Vehicle	0.01 MB	0.03 MB	0.3 MB	0.003 GB
Total	1.8 GB	7.0 GB	56.0 GB	520.6 GB