Skip to main content

GeoWave bindings for Python3

Project description


This project aims to provide Python classes that allow users to interact with a GeoWave data store using the same workflows that are available in the programmatic Java API.


  • Python >=3,<=3.7
  • A virtualenv with requirements.txt installed
  • A running GeoWave Java Gateway

Installation From Source

  • Clone GeoWave: git clone
  • Navigate to python directory: cd geowave/python/src/main/python
  • Set up virtualenv: virtualenv -p python3 venv
  • Activate virtualenv: source venv/bin/activate
  • Install requirements: pip install -r requirements.txt


In order to use pygw, you must have an instance of GeoWave Py4J Java Gateway Server running. The gateway can be started by using the GeoWave command geowave util python rungateway.

You can then import pygw classes into your Python environment.


The following is an example of how pygw might be used to write and query some feature data:

from datetime import datetime

from shapely.geometry import Point

from import DataStoreFactory
from import RocksDBOptions
from pygw.geotools import SimpleFeatureBuilder
from pygw.geotools import SimpleFeatureTypeBuilder
from pygw.geotools import AttributeDescriptor
from pygw.geotools import FeatureDataAdapter
from pygw.index import SpatialIndexBuilder
from pygw.query import VectorQueryBuilder
from pygw.query import VectorAggregationQueryBuilder

# Create a RocksDB data store
options = RocksDBOptions()
# NOTE: Directory is relative to the JVM working directory.
datastore = DataStoreFactory.create_data_store(options)

# Create a point feature type
point_type_builder = SimpleFeatureTypeBuilder()
point_type = point_type_builder.build_feature_type()

# Create a builder for this feature type
point_feature_builder = SimpleFeatureBuilder(point_type)

# Create an adapter for point type
point_type_adapter = FeatureDataAdapter(point_type)

# Create a Spatial Index
index = SpatialIndexBuilder().create_index()

# Registering the point adapter with the spatial index to your datastore
datastore.add_type(point_type_adapter, index)

# Creating a writer to ingest data
writer = datastore.create_writer(point_type_adapter.get_type_name())

# Write some features to the data store
point_feature_builder.set_attr("the_geom", Point(1, 1))

point_feature_builder.set_attr("the_geom", Point(5, 5))

point_feature_builder.set_attr("the_geom", Point(-5, -5))

# Close the writer

# Query the data (with no constraints)
query = VectorQueryBuilder().build()
results = datastore.query(query)
for feature in results:

# Perform a count aggregation on the data (with a CQL constraint)
aggregation_query_builder = VectorAggregationQueryBuilder()
constraints = aggregation_query_builder.constraints_factory().cql_constraints("BBOX(the_geom, 0.5, 0.5, 5.5, 5.5)")
count = datastore.aggregate(

Dev Notes:

Building a distributable wheel

To build a wheel file for pygw, simply execute the command python bdist_wheel --python-tag=py3 under the active virtual environment. This will create a distributable wheel under the dist directory.

Building API documentation

This project has been documented using Python docstrings. These can be used to generate full API documentation in HTML form. To generate the documentation, perform the following steps:

  • Ensure that the GeoWave Py4J Java Gateway Server is running: geowave util python rungateway
  • Generate documentation: pdoc --html pygw

Note: This command requires that the python virtual environment is active and that the pygw requirements have been installed. This will generate API documentation in the html/pygw directory.

Submodule descriptions

In general each submodule tries to mimic the behavior of the GeoWave Java API. If there is ever any question about how something should be done with the Python bindings, the answer is most likely the same as how it is done in Java. The difference being that function names use underscores instead of camel case as is the convention in Java. For example if the Java version of a class has a function getName(), the Python variant would be get_name().

The main difference between the two APIs is how the modules are laid out. The Python bindings use a simplified module structure to avoid bringing in all the unnecessary complexity of the Java packages that the Java variants belong to.


The config module includes a singleton object of type GeoWaveConfiguration called gw_config that handles all communication between python and the Py4J Java Gateway. The module includes several shortcut objects to make accessing the gateway more convenient. These include:

  • java_gateway Py4J Gateway Object
  • java_pkg: Shortcut for java_gateway.jvm. Can be used to construct JVM objects like
  • geowave_pkg: Similar to java_pkg, serves as a shortcut for
  • reflection_util: Direct access to the Py4J reflection utility.

These objects can be imported directly using from pygw.config import <object_name>.

NOTE: the GeoWaveConfiguration has an init() method. This is INTENTIONALLY not an __init__ method. Initialization is attempted when the configuration is imported.


The base module includes common classes that are used by other modules. This includes the base GeoWaveObject class that serves as a python wrapper for a java reference. It also includes a type_conversions submodule that can be used to convert Python types to Java types that are commonly used in GeoWave.


The geotools module contains classes that wrap the functionality of geotools SimpleFeatures and SimpleFeatureTypes. These classes can be used to create feature types, features, and data adapters based on simple features.


The index module contains classes that are used in creating spatial and spatial/temporal indices.


The query module contains classes that are used in constructing queries and their constraints.


The store module contains classes that can be used to establish connections to the various GeoWave backends. Each store type has a submodule which contains a class that can be used to connect to that store type. For example from import AccumuloOptions. The DataStore object can be constructed by passing the options object to the DataStoreFactory.create_data_store(<options>) method.

This exposes a function called print_obj that can be used to help with debugging raw java objects. It will print information about the object in question on both the Python side and on the Java server side. There's a verbose flag that will give you more information about the object in question.


  • j_-prefixed notation : Java reference variables are prefixed with j_ in order to distinguish them from Python variables

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pygw-2.0.1.tar.gz (64.4 kB view hashes)

Uploaded source

Built Distribution

pygw-2.0.1-py3-none-any.whl (127.5 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page