2.1 Release notes | ArcGIS GeoAnalytics Engine

2.1.0 Release notes

ArcGIS GeoAnalytics Engine 2.1.0 introduces new GeoEnrichment capabilities and raster functionality, among other new features and enhancements.

Added support for GeoEnrichment

GeoEnrichment is the process of enhancing your spatial data by appending demographic attributes to polygon areas, describing the people and places within those areas. You can perform GeoEnrichment-related operations in GeoAnalytics Engine using two new analysis tools introduced at version 2.1.0:

GeoEnrichVariables—Returns the complete list of available GeoEnrichment variables for a specified dataset. This tool allows you to inspect available variables and choose those that are appropriate for your analysis before performing GeoEnrichment.
GeoEnrich—Enriches input polygon areas with demographic variables. These variables are calculated for each area using the input GeoEnrichment dataset.

Like all GeoAnalytics Engine functionality, GeoEnrichment is Spark-native and fully scalable. The new GeoEnrichment capabilities complement the existing geocoding and network analysis capabilities within GeoAnalytics Engine and allow you to perform end-to-end workflows from geocoding, to creating service areas, to enriching the service areas with valuable demographics, to subsequent modeling and analysis of patterns across datasets.

A table of geocoded addresses followed by the same table with geoenrichment variables appended. — GeoAnalytics Engine allows you to go from address strings to enriched areas of interest in a single workflow.

This release of GeoAnalytics Engine also includes a new dataset: GeoEnrichment Essentials for ArcGIS GeoAnalytics Engine. This dataset provides a focused set of variables for common GeoEnrichment workflows, allowing you to get started out of the box with the GeoEnrich and GeoEnrichVariables tools. For more information see the core concept topic on GeoEnrichment and the GeoEnrichment dataset setup instructions.

New raster features and enhancements

GeoAnalytics Engine 2.1.0 introduces a new raster tool, Composite Bands, which creates a single output raster by stacking selected bands from one or more input rasters into a unified multiband raster. For example, the tool could be used to combine bands of satellite data stored in separate files into a single raster to create a color composite. In another example, the tool could be used for change detection analyses across multiple time steps. For instance, you could create a single composite dataset with two years of land cover data and then use RT_Apply to quantify land cover change.

Composite bands workflow diagram — The Composite Bands tool could be used to combine raster bands stored in separate files into a single RGB composite.

This release also enhances the Zonal Statistics tool to add support for calculating statistics within square or hexagon bins. As an alternative to using zone polygons that you provide, you can have the tool generate a bin coverage for the spatial extent of your rasters and calculate zonal statistics within each bin.

Version 2.1.0 also includes a new raster function, RT_IsFullyMasked, which offers a simple way to check whether or not all pixels in a given raster are marked as NoData (i.e., masked). NoData pixels will not be used for analysis.

Other new features and enhancements

Starting with GeoAnalytics Engine 2.1.0 the Find Point Clusters tool supports using geodetic distance with the DBSCAN algorithm. Geodetic distance calculations can be especially useful if your input points are not in a projected coordinate system, if your search distance is relatively large, or if you expect to find clusters that span across the antimeridian. For more information see Distance calculations.

This release also includes a new Python function, add_data_licenses(), which allows you to license ArcGIS StreetMap Premium data for use with GeoAnalytics Engine within a running Spark session. This function removes the previous requirement of staging data licenses on your Spark cluster prior to starting Spark, which can be difficult in some cloud environments. For more information, see add licenses using add_data_licenses().

GeoAnalytics Engine 2.1.0 also enhances the register_gis() function to allow registering a GIS with an API key using the apikey parameter. This new approach to registering a GIS is now supported in addition to the already-supported options of using a username and password or a PKI certificate and password. This release also adds the trust_cert_file parameter to register_gis() which can be used to trust ArcGIS Enterprise deployments that use self-signed TLS certificates, or TLS certificates with a root certificate authority not in Java's default trust store.

Spark runtime support

As with every release, version 2.1.0 adds support for new Apache Spark versions and related cloud runtimes. This release includes new compatibility with Databricks 18.2 and AWS EMR 7.13.

This release also resolves a known compatibility issue with Databricks Standard Access Mode and Shared Compute in Databricks runtimes 17.3, 18.0, and 18.1. Please see Databricks compute policy and access mode support for more information.