Skip to content

Releases: opera-adt/dist-s1

dist-s1 v0.0.6

20 Feb 22:44
84cee26
Compare
Choose a tag to compare

Fixed

  • Issues with test_main.py related to where tmp directory was being created (solution, ensure tmp is made explicitly relative to the test directory as in test_workflows.py).
  • All dependencies within virtual environment are back to conda-forge from PyPI.
  • Product directory parameter is now correctly writing to the specified directory (fixes #37).
  • Fixed the CLI test (bug). The runconfig instance will have different product paths than the one created via the CLI because the product paths have the processing time in them, and that is different depending on when the runconfig object is created in the test and within the CLI-run test.

Added

  • Added a n_workers_for_despeckling argument to the RunConfigData model, CLI, and relevant processing functions.
  • A test to ensure that the product directory is being correctly created and used within runconfig (added to test_main.py).

dist-s1 v0.0.5

20 Feb 17:06
759c504
Compare
Choose a tag to compare

Fixed

  • CLI issues with bucket/prefix for S3 upload (resolves #32).
  • Included __main__.py testing for the SAS entrypoint of the CLI; uses the cropped dataset to test the workflow.
  • Includes dist-s1 run_sas testing and golden dataset comparision.
  • Updates to README regarding GPU environment setup.

dist-s1 v0.0.4

14 Feb 18:56
1756e2f
Compare
Choose a tag to compare

Added

  • Minimum working example of generation fo the product via dist-s1 run
  • Integration of dist-s1-enumerator to identify/localize the inputs from MGRS tile ID, post-date, and track number
  • Notebooks and examples to run end-to-end workflows as well as Science Application Software (SAS) workflows
  • Docker image with nvidia compatibility (fixes the cuda version at 11.8)
  • Download and application of the water mask (can specify a path or request it to generate from UMD GLAD LCLUC data).
  • Extensive instructions in the README for various use-case scenarios.
  • Golden dataset test for SAS workflow
  • Allow user to specify bucket/prefix for S3 upload - makes library compatible with Hyp3.
  • Ensure Earthdata credentials are provided in ~/.netrc and allow for them to be passed as suitable evnironment variables.
  • Create a GPU compatible docker image.
  • Ensures pyyaml is in the environment (used for serialization of runconfig).
  • Update equality testing for DIST-S1 product comparison.

Fixed

  • CLI issues with hyp3

Changed

  • Pyproject.toml file to handle ruff

dist-s1 v0.0.3

23 Nov 02:53
f596dbd
Compare
Choose a tag to compare

The OPERA Interface Delivery

Added

  • Python 3.13 support
  • Updated dockerimage to ensure on login the conda environment is activated
  • Instructions in the README for OPERA delivery.
  • A .Dockerignore file to remove extraneous files from the docker image
  • Allow /home/ops directory in Docker image to be open to all users

dist-s1 v0.0.2

22 Nov 03:23
12a9ca1
Compare
Choose a tag to compare

Added

  • Pypi delivery workflow
  • Entrypoint for CLI to localize data via internet (the SAS workflow is assumed not to have internet access)
  • Data models for output data and product naming conventions
  • Ensures output products follow the product and the tif layers follow the expected naming conventions
    • Provides testing/validation of the structure (via tmp directories)

Changed

  • CLI entrypoints now utilize dist-s1 run_sas and dist-s1 run rathern than just dist-s1.
    • The dist-s1 run_sas is the primary entrypoint for Science Application Software (SAS) for SDS operations.
    • The dist-s1 run is the simplified entrypoint for external users, allowing for the localization of data from publicly available data sources.

dist-s1 v0.0.1

20 Nov 17:52
988f7d3
Compare
Choose a tag to compare

Added

  • Initial internal release of the DIST-S1 project. Test github release workflow