[SEDONA-738] Add moran i autocorrelation.

[Imbruced]

Did you read the Contributor Guide?

• Yes

Is this PR related to a ticket?

SEDONA-738

What changes were proposed in this PR?

MoranI index

How was this patch tested?

Unit tests

Did this PR include necessary documentation updates?

Yes

[Imbruced]

I need to add docs

[Imbruced]

and update the description of the MR

[james-willis]

Thanks for this Pawel.

I think we need an ST Function for this. Currently it is asymmetric with the other stats functions in this regard.

I also believe that we should deprecate these function call interfaces for geostats functions and tell users to use the ST functions instead. If we go this direction I would advise against making python and public scala functions for morans I (other than ST functions for the df interfaces of course)

[jiayuasu]

@james-willis Is your SQL function framework able to support Pawel's Moran'I case as well? Maybe you can point him to the right place for this?

[james-willis]

Sure. We should be able to wrap the function in this PR with an implementation of a PhysicalFunction.

Here are the existing ones.

Here is the PR that added SQL support to geostats

[Copilot]

Pull Request Overview

This PR introduces Moran’s I spatial autocorrelation to the Scala, Java, and Python APIs, including implementations, wrappers, tests, and documentation.

• Added Scala implementation and Java result container for Moran’s I
• Exposed Moran’s I in Python with wrapper and updated weighting functions
• Added unit tests (Scala and Python) and extended SQL API documentation

Reviewed Changes

Copilot reviewed 12 out of 13 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
spark/common/src/main/scala/org/apache/sedona/stats/autocorelation/Moran.scala	Core Scala implementation of Moran’s I
spark/common/src/main/java/org/apache/sedona/stats/autocorrelation/MoranResult.java	Java container class for result values
spark/common/src/main/scala/org/apache/sedona/stats/Weighting.scala	Added addDistanceBandColumnPython bridge method
spark/common/src/test/scala/org/apache/sedona/stats/autocorellation/MoranTest.scala	Scala tests for positive, negative, and zero cases
spark/common/src/test/scala/org/apache/sedona/stats/autocorellation/AutoCorrelationFixtures.scala	Test fixtures for auto-correlation scenarios
python/sedona/spark/stats/autocorrelation/moran.py	Python wrapper for Moran’s I
python/tests/stats/test_moran.py	Python integration tests for Moran’s I
python/sedona/spark/stats/weighting.py	Updated Python weighting functions to call wrapper
python/sedona/spark/stats/hotspot_detection/getis_ord.py	Fixed DataFrame wrapping in g_local call
python/sedona/spark/register/java_libs.py	Registered Moran in JVM libs enum
python/sedona/spark/stats/autocorrelation/init.py	Initialized Python autocorrelation package
docs/api/stats/sql.md	Documented MoranI algorithm and usage

Comments suppressed due to low confidence (3)

spark/common/src/test/scala/org/apache/sedona/stats/autocorellation/MoranTest.scala:1

• The package name 'autocorellation' is misspelled and does not match the implementation package 'autocorrelation'. Please rename the directory and package declaration to 'autocorrelation' so the tests compile correctly.

/*

docs/api/stats/sql.md:142

• [nitpick] Minor grammatical issue: remove 'the' before '1' to read 'close to 1', and consider rephrasing to improve clarity.

location and non-spatial attribute. When the value is close to the 1 it

docs/api/stats/sql.md:146

• Grammar correction: change 'has' to 'have' since 'values' is plural.

correlation. Negative correlation means that close values has dissimilar values.

[jiayuasu]

@Imbruced I will merge this for now but the doc needs significant improvement. I believe we need some comprehensive end-to-end examples of each Stat function

[james-willis]

Filed a ticket for ST function: #2115