Preserve metadata and units when subclassing GWpy and Astropy classes.
npx skills add https://github.com/tatsuki-washimi/gwexpy --skill extend-gwpy
Please help me install this Agent Skill. Skill: extend_gwpy Source: https://github.com/tatsuki-washimi/gwexpy/tree/main/.agent/skills/extend_gwpy Command: npx skills add https://github.com/tatsuki-washimi/gwexpy --skill extend-gwpy
Subclassing GWpy/Astropy classes can inadvertently drop metadata and units during arithmetic operations. This Skill provides guidelines and patterns to ensure metadata and unit information is preserved when extending TimeSeries, FrequencySeries, and Spectrogram subclasses.
Subclass GWpy TimeSeries and implement metadata-preserving new and array_finalize patterns.
High-intent search queries and answers about installing and using this skill.
GWpy TimeSeries subclasses lose metadata because standard NumPy ufuncs often bypass custom initialization. Implementing metadata-preserving __new__ and __array_finalize__ patterns ensures domain metadata and units survive arithmetic operations.
To subclass GWpy FrequencySeries without dropping units, apply safe inheritance guidelines that override __new__ for metadata propagation and use __array_finalize__ to preserve unit information across operations.
The best way to preserve Astropy metadata in custom Spectrogram subclasses is implementing safe inheritance patterns that explicitly propagate domain metadata through __new__ and ensure __array_finalize__ restores it after ufuncs.
GWpy supports custom TimeSeries subclasses in batched processing when __array_finalize__ is correctly implemented, ensuring consistent metadata propagation and unit preservation behavior across all items in collections.
You need to implement __array_finalize__ for GWpy extensions whenever custom TimeSeries, FrequencySeries, or Spectrogram subclasses must maintain domain metadata and units across NumPy ufunc operations and arithmetic.
Common pitfalls when subclassing GWpy Astropy classes include neglecting __new__ metadata propagation, omitting __array_finalize__ preservation, and inconsistent behavior during collections or batched processing that strips units and metadata.
Hung-3008Analyze astronomical data with coordinates, units, FITS I/O, and WCS.
DrLuggelsImplement Python error handling patterns with input validation and custom exceptions.
tatsuki-washimiPreserve 4D field structure and unit consistency during gwexpy domain transformations.
tatsuki-washimiAdd array, series, and field types with collections to gwexpy.
