Apache Sedona

Today I'm releasing the first experimental build of CereusDB, a WebAssembly version of SedonaDB that runs in browsers: https://lnkd.in/daCDdAMv Recently I dug a bit deeper into Apache Sedona when I saw that they also offer SedonaDB, "an open-source single-node analytical database engine with geospatial as a first-class citizen". SedonaDB is built in Rust upon DataFusion, which theoretically enables Wasm builds via Emscripten. It relies heavily on the usual library suspects like GEOS, PROJ, S2 and GDAL, which makes compilation for Wasm targets a little more complicated. I used Codex with GPT-5.4 high/xhigh to iterate on an implementation, and after hitting some dead ends, it finally succeeded. Including all features wouldn't be a small library, so I decided to have different build which contain functional subsets (uncompressed / gzipped / brotli size in MB in the brackets): - minimal: Core + geo + GEOS + spatial joins / ST_KNN (21.3 / 6.2 / 4.0) - standard: minimal + PROJ / ST_Transform (40.6 / 10.8 / 6.0) - global: standard + opt-in S2 geography kernels (42.9 / 11.8 / 6.7) - full: global + GDAL-backed raster ingestion and the full RS_* functions (49.5 / 14.4 / 8.5) With the right headers, it can easily be cached in the browsers so it's only downloaded once and not on every page load. Let me know what you think, happy for any feedback!

Our monthly office hour is coming up! We'll share updates on the upcoming Sedona 1.9.0 release and the out-of-core spatial join in SedonaDB 0.3.0. Bring any questions you have. We'd love to hear how you're using Apache Sedona in your spatial workflows. 📆 Tuesday, April 7th at 8AMPT / 11AM ET ➡️ https://lnkd.in/gQ5BR-KY

💥 6 Million Trips. 156K Zones. 512 MB of RAM. A Rust Spatial Database That Makes It Possible

We're chuffed to announce Apache SedonaDB 0.3.0! This release features a rewritten join that supports larger-than-memory spatial/KNN joins courtesy of Kristin Cowalcijk, 36 new functions, item-level CRSes, parameterized SQL queries, GDAL/pyogrio-based writes for FlatGeoBuf, Shapefile, and GeoPackage, GDAL/sf based reads in R, and the beginnings of an R DataFrame API. Find us in Python or R to get started or check out the release post (in the comments).

We’re back with our monthly office hour session, and there’s a lot to share this time 👇 What’s new in Apache Sedona 1.9.0 • Cloud Optimized GeoTIFF writer support • New Bing tile function • More comprehensive CRS format support in ST_Transform • Support for Apache Spark 4.1 SedonaDB 0.3 release • Out-of-core spatial join • Item-level CRS support If you’re building or scaling geospatial workloads, join us to get the latest updates and bring your questions. 📆 Tuesday, March 3 at 8AM PT 🔗 Save the date: https://lnkd.in/gygDqVb8

The Wherobots team in collaboration with The Apache Software Foundation parquet, Apache Iceberg and Apache Sedona communities led the inclusion of spatial type (specifically GEOMETRY and GEOGRAPHY) as first-class logical types within the Apache Parquet columnar storage format. Spatial data types are quite essential to representing data collected from the real physical world, such as vehicle / robot movement, weather forecast, farmland, buildings details, and generally anything happening on the surface of the earth. By moving away from treating spatial data as opaque binary blobs, Parquet now enables query engines to perform efficient data skipping through the use of spatial statistics, such as bounding boxes stored at the row-group level. This integration supports standardized Well-Known Binary (WKB) encoding and includes built-in metadata for Coordinate Reference Systems (CRS), ensuring seamless interoperability across different data processing engines and lakehouse architectures like Apache Iceberg. Ultimately, this update eliminates the need for specialized spatial sidecar files, allowing spatial workloads to benefit from the same performance optimizations and scalability as standard tabular data in the lakehouse architecture. Check out the article on the official Apache Parquet blog: https://lnkd.in/g-edwsKu

Join us next week for our monthly office hour session! We'll introduce features from several recent releases across the Apache Sedona ecosystem, including: - Apache Sedona 1.8.1 - SedonaDB 0.2.0 - SpatialBench 0.1.0. ⁠⁠ Bring your questions and save the date: https://lnkd.in/gQ5BR-KY

I'm a bit late to the #ModernGeospatial party, but I recently got around to kicking the tires of some relative newcomers to the geospatial scene- namely DuckDB, SedonaDB, and Geoparquet. I wrote two scripts that you can access at the link below that do the same thing- download Cook County, IL building footprints as a GeoPackage from Overture Maps. One uses DuckDB, the other uses SedonaDB. What excites me about these tools is their flexibility and speed. We used to have to download massive files, then unzip them, then transform them in a desktop GIS program. No more. Write some SQL, and get exactly what you want in one step- quickly too. Shoutout to Matt Forrest, Qiusheng Wu, and many others for so consistently putting out such great content on all things modern geospatial. https://lnkd.in/gK7B_eT4