adjusting links to enable pixel, wavelength linking in specviz2d

CHANGES.rst

@@ -114,6 +114,8 @@ Specviz
 Specviz2d
 ^^^^^^^^^
 
+- Fix subset linking/displaying between pixel/wavelength in Specviz2d viewers. [#2736]
+
 4.1.1 (2025-01-31)
 ==================
 
@@ -747,6 +749,8 @@ Specviz
   display label for the y axis in spectral viewer. Previously it was hard coded
   to always display ``flux density`` no matter the input unit. [#2703]
 
+Specviz2d
+^^^^^^^^^
 
 3.8.2 (2024-02-23)
 ==================

jdaviz/app.py

@@ -766,8 +766,22 @@ def _link_new_data(self, reference_data=None, data_to_be_linked=None):
                  linked_data.ndim < 3 and  # Cube linking requires special logic. See below
                  ref_data.ndim < 3)
               ):
-            links = [LinkSame(linked_data.components[0], ref_data.components[0]),
-                     LinkSame(linked_data.components[1], ref_data.components[1])]
+            if self.config == 'specviz2d':
+                links = []
+                if linked_data.ndim == 2:
+                    # extracted image added to data collection
+                    ref_wavelength_component = ref_data.components[1]
+                else:
+                    # extracted spectrum added to data collection
+                    ref_wavelength_component = ref_data.components[3]
+                    links += [LinkSameWithUnits(linked_data.components[0], ref_data.components[1])]
+
+                links += [LinkSameWithUnits(linked_data.components[0], ref_data.components[0]),
+                          LinkSameWithUnits(linked_data.components[1], ref_wavelength_component)]
+            else:
+                links = [LinkSame(linked_data.components[0], ref_data.components[0]),
+                         LinkSame(linked_data.components[1], ref_data.components[1])]
+
             dc.add_link(links)
             return
 

jdaviz/tests/test_subsets.py

@@ -1,9 +1,11 @@
 import numpy as np
 import pytest
 from astropy import units as u
+from astropy.wcs import WCS
 from astropy.tests.helper import assert_quantity_allclose
 from astropy.utils.data import get_pkg_data_filename
-from glue.core.roi import CircularROI, CircularAnnulusROI, EllipticalROI, RectangularROI
+from glue.core.roi import CircularROI, CircularAnnulusROI, EllipticalROI, RectangularROI, XRangeROI
+
 from glue.core.subset_group import GroupedSubset
 from regions import (PixCoord, CirclePixelRegion, CircleSkyRegion, RectanglePixelRegion,
                      EllipsePixelRegion, CircleAnnulusPixelRegion)
@@ -740,6 +742,104 @@ def test_only_overlapping_in_specviz2d(specviz2d_helper, mos_spectrum2d):
     assert reg[1].lower.value == 7600 and reg[1].upper.value == 7800
 
 
+def test_draw2d_linking_specviz2d(specviz2d_helper):
+    # custom test data to predict values for different viewers
+    header = {
+              'WCSAXES': 2,
+              'CRPIX1': 0.0, 'CRPIX2': 8.5,
+              'CDELT1': 1E-06, 'CDELT2': 7.5E-05,
+              'CUNIT1': 'm', 'CUNIT2': 'deg',
+              'CTYPE1': 'WAVE', 'CTYPE2': 'OFFSET',
+              'CRVAL1': 0.0, 'CRVAL2': 5.0,
+              'RADESYS': 'ICRS', 'SPECSYS': 'BARYCENT'}
+    wcs = WCS(header)
+
+    x_values = np.linspace(0, 10, 128)
+    y_values = np.linspace(0, 5, 256)
+
+    # Create a continuous 2D
+    data = np.sin(x_values[:, np.newaxis]) * np.cos(y_values) * u.one
+    spectrum_data = Spectrum1D(data, wcs=wcs, meta=header)
+
+    specviz2d_helper.load_data(spectrum_2d=spectrum_data)
+    viewer_1d = specviz2d_helper.app.get_viewer(
+        specviz2d_helper._default_spectrum_viewer_reference_name)
+    viewer_2d = specviz2d_helper.app.get_viewer(
+        specviz2d_helper._default_spectrum_2d_viewer_reference_name)
+
+    # create subset in 2d viewer,  want data in 1d viewer
+    viewer_2d.apply_roi(XRangeROI(60, 80))
+    subset_drawn_2d = viewer_1d.native_marks[-1]
+
+    # get x and y components to compute subset mask
+    y1 = subset_drawn_2d.y
+    x1 = subset_drawn_2d.x
+
+    subset_highlighted_region1 = x1[np.isfinite(y1)]
+    min_value_subset1 = np.min(subset_highlighted_region1)
+    max_value_subset1 = np.max(subset_highlighted_region1)
+
+    tolerance1 = 1e-6
+    expected_min1 = 6.e-5
+    expected_max1 = 8.e-05
+
+    assert np.allclose(min_value_subset1, expected_min1, atol=tolerance1)
+    assert np.allclose(max_value_subset1, expected_max1, atol=tolerance1)
+
+
+'''
+def test_draw1d_linking_specviz2d(specviz2d_helper):
+    # custom test data to predict values for different viewers
+    header = {
+            'WCSAXES': 2,
+
+            'CRPIX1': 0.0, 'CRPIX2': 8.5,
+            'CDELT1': 1E-06, 'CDELT2': 7.5E-05,
+            'CUNIT1': 'm', 'CUNIT2': 'deg',
+            'CTYPE1': 'WAVE', 'CTYPE2': 'OFFSET',
+            'CRVAL1': 0.0, 'CRVAL2': 5.0,
+            'RADESYS': 'ICRS', 'SPECSYS': 'BARYCENT'}
+    wcs = WCS(header)
+
+    x_values = np.linspace(0, 10, 128)
+    y_values = np.linspace(0, 5, 256)
+
+    # Create a continuous 2D image
+    data = np.sin(x_values[:, np.newaxis]) * np.cos(y_values) * u.Jy
+    spectrum_data = Spectrum1D(data, wcs=wcs, meta=header)
+
+    specviz2d_helper.load_data(spectrum_2d=spectrum_data)
+    se = specviz2d_helper.plugins['Spectral Extraction']
+    se.export_extract()
+
+    viewer_1d = specviz2d_helper.app.get_viewer(
+        specviz2d_helper._default_spectrum_viewer_reference_name)
+    viewer_2d = specviz2d_helper.app.get_viewer(
+        specviz2d_helper._default_spectrum_2d_viewer_reference_name)
+
+    # subset drawn in 1d viewer, want data in 2d viewer
+    spec_reg = SpectralRegion(0. * u.m, 10000000 * u.m)
+    spec_reg = SpectralRegion(0.0001 * u.m, .0002 * u.m)
+    st = specviz2d_helper.plugins['Subset Tools']
+    st.import_region(spec_reg)
+
+    subset_drawn_1d = viewer_2d.native_marks[-1]
+
+    subset_highlighted_region2 = np.atleast_1d(np.nonzero(subset_drawn_1d))[1]
+
+    # Get the start and stop indices
+    min_value_subset2 = np.min(subset_highlighted_region2)
+    max_value_subset2 = np.max(subset_highlighted_region2)
+
+    tolerance2 = 2
+    expected_min2 = 209
+    expected_max2 = 300
+
+    assert np.allclose(min_value_subset2, expected_min2, atol=tolerance2)
+    assert np.allclose(max_value_subset2, expected_max2, atol=tolerance2)
+'''
+
+
 def test_multi_mask_subset(specviz_helper, spectrum1d):
     specviz_helper.load_data(spectrum1d)
     subset_plugin = specviz_helper.plugins['Subset Tools']._obj