apache / apache/sedona-db

rust/sedona-spatial-join: Make spatial partitioners wraparound aware or introduce a geography partitioner

Open
#800 2 comments 0 reactions 0 assignees View on GitHub
Dominant language
Rust
Stars
503
Forks
61
Avg merge
2d 4h
Merged PRs (30d)
90

Description

In https://github.com/apache/sedona-db/pull/799 we enable wraparound rectangles in the index and analysis statistics; however, this means that some GeoStatistics might have wraparound bounds which will cause the partitioner to error. This should be safe in the sense that wraparound bounds error and aren't blindly interpreted, and the workaround is to set the memory limit to unlimited (i.e., disable the out of core join). To make this work we have to extend wraparound intervals to the right or left of -180 or 180 to make the partition bounds finite, or possibly something else I haven't considered.

Contributor guide

Open the contributing guide

Research direction

Start by reviewing PR #799 and the spatial-join partitioner and GeoStatistics behavior described here. Reproduce the failure with wraparound bounds, then determine whether extending intervals or introducing a geography partitioner is the appropriate approach; done means wraparound statistics no longer cause partitioning to error while preserving safe handling of invalid bounds.

Written by the indexing model from the issue text.

Assessment

Tech stack
rust
Domain
databases
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.