Unclear what the real underlying issue is. Code comments give a
possible explanation.

…otlib_standalone_compat

fix: make standalone work with matplotlib >= 3.9.0

…ld them

…cope#2934)

* support exporting a plot from an unopened/inactive plugin
* prevent clearing plot selection (under most conditions)

Sync notebook github links in HTML docs to the git commit used to build them

* BUG: Specviz to respect local_path and timeout when provided

…traction.py

Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>

---------

Co-authored-by: Katherine Carver <kcarver@osx-summer2401.stsci.edu>
Co-authored-by: P. L. Lim <2090236+pllim@users.noreply.github.com>
Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>
Co-authored-by: Ricky O'Steen <39831871+rosteen@users.noreply.github.com>

…scope#2767

…orientation-dropdown

Exclude subsets from WCS-only layer filter

* Use normalize_filename and overwrite_warn for plugin plots

* Clear overwrite warning overlay after confirming

* Changelog

* Codestyle

* Markers: Save to CSV

* Mark block as uncovered for coverage

* Remove unused entries from _dict

* Update CHANGES.rst

Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>

---------

Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>

…escope#2941)

* Create private method for importing spectral region objects

* Address review comments

* Add tests for list of modes and single SpectralRegion

* Fix style

…scope#3096)

Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>

…-test

Fix for Imviz parser romandeps test after update to defaults

Changelog update for release of v3.10.3

…extract

Bugfix for flux unit conversion in spectral cubes

* UI functioning unit conversion, need to reconcile tests

* update syntax, remove unnecessary logic

* tests now pass, API functionality improved

* add additional equivalencies for coords info

* first pass at disabling specviz translations

* Remove dev flag, consolidate untranslatable units, multi-config support, styling changes

* Refactor moment map unit conversion
to not require spectrum, so nothing to centralize now.

Fix devdeps

Simplify flux_conversion logic
and other minor fixes. Add flux conversion tests.

Clarify which one should be reverted
in the future

* Remove SB/flux conversion for moment map
because moment map always in SB

* add os import

* remove duplicate import from rebase

* reconcile tests, change hints/docs, handle dimensionless case

* emit GlobalDisplayUnitChanged msg when toggling SB<>flux

* resolve most tests

* temporarily remove translation message

* add comments, reenable global display message, resolve tests

* force sb unit for line flux test

* add comments, clarify naming, add case handling for args

* remove args and loop, replaced with msg

* consolidate 2 translator functions into 1, consolidate _translate and _on_flux_unit_changed

* address test failures

* update marks.py equivalencies, resolving UnitConversion errors

* make sure not only translation equivalencies are in marks.py but also conversions

* add test coverage

* use astropy units, small logic tweaks, cache untranslatable units

* technical review changes

* add warning message in tray if PIXAR_SR not present in FITS header

* second pass at technical changes

* move flux_unit checks to beginning of _on_flux_unit_changed

* change _unit to flux.unit

---------

Co-authored-by: P. L. Lim <2090236+pllim@users.noreply.github.com>
Co-authored-by: Kyle Conroy <kyleconroy@gmail.com>

…telescope#3101)

* TST: Also run slow test in non-dev env

* BUG: Fix table I/O in Catalog Search

* ID as string always

* Force ID to be always str even from file

…ope#3114)

* MNT: Remove unnecessary astropy version checks

* TST: Update model doctest

…pacetelescope#3113)

* Enabling model fitting results to convert between flux and surface brightness

* Switch to using selected_obj, clean up pixel scale retrieval

* Changelog

* Don't store pixel scale factor

* Fix missing variable

* Update test value due to updated pixel scale factor

* generalize use of term "spectrum"
* move code to allow importing slice mixins without circular imports
* is_slice_indicator filter for extract results instead of relying on a spectrum viewer existing and being set
* attach marks to slice_indicator_viewers[0] instead of spectrum_viewer (temporarily to just the first instance, later we'll make this more flexible to allow any number of slice indicator viewers)
* remove reliance on loaded_flux_cube
* rename (internal) spectral_cube -> cube
* generalize logic for setting disabled_msg
* generalize internal marks names
* generalize some internal traitlet names
* update disabled message to specify cube
* update tests for renamed marks/method
* generalize logic within live-preview and extract
* allow skipping wavelength-dependence
* hide aperture masking method for entire cube and no bg
* fix (and allow disabling) bg export
* fix updating live preview on change to subset
* skip updating marks if not tray instance

MNT: Remove unnecessary astropy check.

…escope#3121)

* remove "Imviz" from "Line Profiles XY" plugin name with deprecation

* changelog entry

spacetelescope#2931)

* Respond to changed unit and display on the UI

* Convert all sb<>sb units listed in the dropdown menu

* Use the default current unit from the cube

* Apply only for relevant configs

* Update current_unit setting logic

* Fix failing tests

* Add tests

* Add a changelog

* Move conversion logic to utils

* Move the conversion to utils

* Pass when unable to convert

* Update tests

* Update based on the feedback

* Only subscribe to cubeviz

* Check for equivalency rather than image dimension

* Refactor conversion logic

Co-authored-by: P. L. Lim <2090236+pllim@users.noreply.github.com>

* Rename unit and broadcast 'sb' as part of msg

* Code style

---------

Co-authored-by: P. L. Lim <2090236+pllim@users.noreply.github.com>

TST: Fix tests downloading data into source dir, MNT: Remove unnecessary astropy check

to write out fitted cube.

This uses astropy/specutils#1051

Cubeviz to use wcs1d-fits from specutils to write out fitted cube

* Avoid checking selected_obj before it's available

* Changelog

* Added a test that fails on main

- Fix sigcube alignment with other cubes [get rid of transpose]
- Add percentile cut option that enables some rough feature isolation without full continuum subtraction
- Protect against cubes w/NaNs

…ions