Releases: opendatacube/odc-stac
Releases Β· opendatacube/odc-stac
v0.3.0
Major Changes from v0.2.4
- No longer depend on
datacube
library- Significantly smaller set of compulsory dependencies, easier to install/deploy
- Using
odc-geo
library instead ofdatacube
forGeoBox
andGeometry
classes - Can load data into rotated pixel planes (Example)
- Arbitrary grouping of STAC items into pixel planes with user supplied grouping methods or group by property name
- Better handling of credentials and other GDAL state in distributed context
- credentials and GDAL environment configuration were part of the global state previously, now global state is removed, so you can access collections with different permissions from the same Dask cluster (for example mixing public and private access).
- Parallelized data loading even when not using Dask
- Progress reporting for non-Dask load with
tqdm
See demo notebook highlighting some of the new features of odc-stac
and features enabled by transition to odc-geo
.
What's Changed
Auto-generated changes follow here
- New example showing how to load DE Africa data by @caitlinadams in #32
- Fix build by @Kirill888 in #33
- Change band name from SLC to SCL by @caitlinadams in #35
- Fixed one remaining instance of SLC by @caitlinadams in #36
- Tweak README a bit (#38) by @Kirill888 in #39
- Tweak docs in stac.load by @Kirill888 in #41
- Starting migration away from datacube dependency by @Kirill888 in #42
- De-duplicate code in _eo3converter by @Kirill888 in #43
- Adding output geobox computation by @Kirill888 in #44
- Low level reader functionality by @Kirill888 in #45
- Starting to add some tests for internal reader by @Kirill888 in #51
- Set floor for stackstac version by @gadomski in #53
- Starting new stac.load version by @Kirill888 in #57
- Implement temporal grouping and populate time axis by @Kirill888 in #58
- Implementing load with pixel fusing by @Kirill888 in #60
- Implement per band resampling and dtype config by @Kirill888 in #61
- Implement Dask version of load by @Kirill888 in #62
- Add back configure_rio and configure_s3_access by @Kirill888 in #63
- Merge develop changes to next by @Kirill888 in #64
- Fix packaging and bump version to 0.3.0a1 by @Kirill888 in #65
- Update sample notebooks and binder environment by @Kirill888 in #66
- Next version: no datacube dependency by @Kirill888 in #67
- Release candidate (0.3.0rc0) by @Kirill888 in #68
- Progress reporting and multi-threaded load by @Kirill888 in #69
- Fix issue #70, release 0.3.0rc2 by @Kirill888 in #71
- Generalize groupby by @Kirill888 in #72
New Contributors
- @caitlinadams made their first contribution in #32
Full Changelog: v0.2.4...v0.3.0
v0.2.4
v0.2.3
Added
- This CHANGELOG
requirements-dev.txt
- Documentation
- Upload built conda environment as an artifact
- Notebook rendering to Github actions, including hash-based artifact checks
- Initial benchmarking tooling, still in progress
Changed
- Moved publishing steps into separate workflows
- Deprecated imports from
odc.index.*
- Removed
.units
attribute from.time
axis for better inter-op with.to_zarr
,.to_netcdf
Fixed
- Handling of STAC Items with only partial
proj
data - Typos in documentation