Skip to content

CHANGELOG.md

Latest commit

 

History

History
84 lines (64 loc) · 4.3 KB

CHANGELOG.md

File metadata and controls

84 lines (64 loc) · 4.3 KB

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project follows Semantic Versioning.

Unreleased

0.2.0 - 2026-03-02

Added

  • query-cutana CLI command for generating Cutana input files from object IDs or coordinates.
  • Canonical spectra compilation support for chunk-level parallel workers via compile-spectra --workers.
  • Datalink spectra compilation mode with compile-spectra --use-datalink.
  • Sphinx + Read the Docs documentation scaffold and user guide pages.

Changed

  • Default spectra chunk size changed to 1000 extensions per output FITS.
  • Spectra compilation now resumes by default and can continue correctly across previous runs with different chunk sizes.
  • Crossmatch now includes point_like_prob and extended_prob in outputs.
  • Credential setup guidance updated to use secure user-local files under ~/.euclidkit.
  • Project license updated to BSD 3-Clause for public release/distribution.

Fixed

  • Crossmatch async mode now attempts result download and writes table output when available.
  • Crossmatch object-id mode logging now reports mode-specific messaging (no radius confusion).
  • Metadata table generation in spectra compilation fixed for mixed chunk/resume scenarios.

0.2.0rc.4 - 2026-02-23

Changed

  • Resume logic for canonical spectra compilation now inspects existing chunk FITS files and skips rows based on actual compiled extension counts.
  • Resume now supports differing historical chunk sizes by continuing from already-compiled row count rather than assuming the current chunk size.
  • Metadata generation now maps chunk/extension assignments from real output files, improving correctness for mixed-size resume cases.

0.2.0rc.3 - 2026-02-22

Changed

  • README cleaned for public release: removed unimplemented feature claims and examples.
  • Added a project changelog file and release entries.

0.2.0rc.2 - 2026-02-22

Added

  • compile-spectra --workers option for chunk-level parallel compilation in canonical (local FITS) mode.
  • Resume behavior for compilation by default when output chunks already exist.
  • Datalink spectrum retrieval path using private astroquery client (__eucliddata.load_data) for improved robustness.
  • compile-spectra --limit option to process only the first N rows for quick tests.
  • Additional CLI tests for datalink mode, resume behavior, and worker forwarding.

Changed

  • Spectrum compilation now logs grouping statistics for FITS-path grouping per chunk.
  • Compiled FITS extension header key standardized to SOURC_ID for consistency with archive datalink FITS headers.
  • Metadata output typing hardened to avoid object-dtype FITS write failures.
  • compile-spectra output directory handling now prefers in-place resume and only prompts on explicit overwrite requests.

Fixed

  • Datalink handling for empty/invalid downloads after server-side 500 responses.
  • Metadata table generation error: chunk_file object dtype caused FITS serialization failures.

0.2.0rc.1 - 2026-02-21

Added

  • New query-cutana CLI command to build Cutana input tables from source tables.
  • Support for IDR field selection (WIDE/DEEP) in Cutana-related archive lookups.
  • Improved async behavior for large crossmatch queries.
  • Initial PyPI publishing workflow via GitHub Actions.

Changed

  • Cutout defaults updated to 15 arcsec.
  • Query logic updated to rely on segmentation_map_id workflows for mosaic product retrieval.
  • Environment-dependent table prefix handling refined (q1/dr1 behavior across PDR/REG/IDR).

Fixed

  • Removed legacy distance-based matching fields and unnecessary DISTANCE(...) projections in cutout query paths.
  • Corrected join strategy to use object-ID joins against MER catalog where applicable.