1.6.0 Release notes
Scala API for tools
This release includes the addition of a Scala API for all tools in GeoAnalytics Engine. You can now drive analysis tools using either Python or Scala syntax while utilizing the same underlying implementation. This complements the Scala API for SQL and TRK functions that was added with version 1.5.0.
Write polygons and linestrings to vector tiles
With GeoAnalytics Engine 1.6.0 you can now write polygon and linestring geometries to vector tiles in addition to points. This allows for hosting and fast visualization of your spatial data, no matter the geometry type. For example, you can aggregate point observations into high-resolution H3 bins and quickly host and visualize the bins in a web map.
Find Co-travelers group mode
This release enhances the Find Co-travelers tool to enable finding groups
of two or more entities co-traveling together using the set method. When group mode is enabled, each record
in the result represents a single co-travel event between a minimum number of entities that you define. For example,
using a dataset of bicycle locations, you can discover common routes for group rides by finding where 5 or more
bicyclists have traveled together for some amount of time.
SQL and TRK functions
This release introduces two new SQL functions: ST_LineInterpolatePoint and ST_LineLocatePoint. ST_LineInterpolatePoint takes a linestring and a value between 0 and 1 and returns a point geometry representing the location at a fractional distance along the linestring. ST_LineLocatePoint is the complementary operation in that it returns the fractional distance along a linestring at which it intersects a point.
SQL and TRK functions have also been enhanced to now support specifying any distance or duration parameter using the
ST_CreateDistance and
ST_CreateDuration functions. This lets you specify a different distance or duration
value for each row in your DataFrame. You can alternatively use a tuple containing a number and unit
(e.g., (5, "meters") or (2, "minutes")) to use the same distance or duration for all rows in your DataFrame.
Performance enhancements
The 1.6.0 release of GeoAnalytics Engine includes a new quantization option for faster plotting of geometries with st.plot. When using quantization, a relatively small amount of resolution is lost in exchange for much better performance when plotting, which may be useful when visualizing intermediate results of your analysis in a PySpark notebook.
This release also includes improved performance of spatial joins and related operations for some cases.
New Spark and cloud runtime support
With every release, support for the latest Spark versions and cloud runtimes is added if possible. In 1.6.0, GeoAnalytics Engine adds support for Spark 3.5.4, 3.5.5 and the latest of the following cloud runtimes:
- Databricks (16.1, 16.2, 16.3, and 16.4)
- AWS EMR (7.4.0, 7.5.0, 7.6.0, 7.7.0, 7.8.0, and 7.9.0)
With the 1.6.0 release of GeoAnalytics Engine, support has been dropped for Spark 3.0.x and 3.1.x, along with any cloud runtimes that depend on these Spark versions.