Abstract
New Features ^^^^^^^^^^^^
• Convolution-based smoothing will now apply a 1D kernel to multi-dimensional fluxes by convolving along the spectral axis only, rather than raising an error. [#885]
• template_comparison now handles astropy.nddata.Variance and astropy.nddata.InverseVariance uncertainties instead of assuming the uncertainty is standard deviation. [#899]
Bug Fixes ^^^^^^^^^
• Speed up JWST s3d loader and reduce memory usage. [#874]
• SpectralRegion can now handle pixels. [#886]
• Fix bug where template_comparison would return the wrong chi2 value. [#872]
Other Changes and Additions ^^^^^^^^^^^^^^^^^^^^^^^^^^^
• fit_lines now makes use of unit support in astropy.modeling. [#891]
• Spectrum1D.with_spectral_units now attempts to fall back on the spectral_axis units if units could not be retrieved from the WCS. [#892]
• ndcube package pin updated to released version (2.0). [#897]
• Minor changes for astropy 5.0 compatibility. [#895]
• Convolution-based smoothing will now apply a 1D kernel to multi-dimensional fluxes by convolving along the spectral axis only, rather than raising an error. [#885]
• template_comparison now handles astropy.nddata.Variance and astropy.nddata.InverseVariance uncertainties instead of assuming the uncertainty is standard deviation. [#899]
Bug Fixes ^^^^^^^^^
• Speed up JWST s3d loader and reduce memory usage. [#874]
• SpectralRegion can now handle pixels. [#886]
• Fix bug where template_comparison would return the wrong chi2 value. [#872]
Other Changes and Additions ^^^^^^^^^^^^^^^^^^^^^^^^^^^
• fit_lines now makes use of unit support in astropy.modeling. [#891]
• Spectrum1D.with_spectral_units now attempts to fall back on the spectral_axis units if units could not be retrieved from the WCS. [#892]
• ndcube package pin updated to released version (2.0). [#897]
• Minor changes for astropy 5.0 compatibility. [#895]
| Original language | English |
|---|---|
| DOIs | |
| Publication status | Published - 23 Nov 2021 |
Fingerprint
Dive into the research topics of 'astropy/specutils: V1.5.0'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver