Plugin plots to viewers

CHANGES.rst

@@ -4,6 +4,9 @@
 New Features
 ------------
 
+- Plots in plugins now include basic zoom/pan tools for Plot Options,
+  Imviz Line Profiles, and Imviz's aperture photometry. [#2498]
+
 Cubeviz
 ^^^^^^^
 

jdaviz/app.vue

@@ -360,7 +360,12 @@ a:active {
   text-decoration: none;
 }
 
-.jdaviz-nested-toolbar {
+.invert, .invert-if-dark.theme--dark {
+    filter: invert(1) saturate(1) brightness(100);
+    color: white;
+}
+
+.jdaviz-nested-toolbar, .plugin-nested-toolbar {
   /* height of nested toolbar to match viewer toolbar height */
   height: 42px;
   margin-right: 4px;
@@ -371,12 +376,7 @@ a:active {
   filter: invert(1) saturate(1) brightness(100);
 }
 
-.invert, .invert-if-dark.theme--dark {
-    filter: invert(1) saturate(1) brightness(100);
-    color: white;
-}
-
-.jdaviz-nested-toolbar .v-btn {
+.jdaviz-nested-toolbar .v-btn, .plugin-nested-toolbar .v-btn {
   height: 42px !important;
   border: none !important;
   min-width: 42px !important;
@@ -394,6 +394,11 @@ a:active {
   background-color: #c75109 !important;
 }
 
+.plugin-nested-toolbar .v-btn--active, .plugin-nested-toolbar .v-btn:focus {
+  /* semi-transparent active color (orange) */
+  background-color: #c7510996 !important;
+}
+
 .v-divider.theme--dark {
   /* make the v-divider standout more */
   border-color: hsla(0,0%,100%,.35) !important;

jdaviz/components/plugin_plot.vue

@@ -1,22 +1,22 @@
 <template>
-  <div class="plugin-plot-component">
-    <v-row style="margin: 0px 0px -8px 0px !important" justify="end">
-      <div style="line-height: 64px; width=32px" class="only-show-in-tray">
+  <div class="plugin-plot-component" style="margin-bottom: 40px">
+    <v-row dense style="margin: 0px 0px -8px 0px !important">
+      <jupyter-widget class='plugin-nested-toolbar' :widget="toolbar"></jupyter-widget>
+      <v-spacer></v-spacer>
+      <div style="line-height: 40px; width=32px" class="only-show-in-tray">
         <j-plugin-popout :popout_button="popout_button"></j-plugin-popout>
       </div>
     </v-row>
 
     <v-row style="margin: -16px 0px 8px 0px !important">
       <jupyter-widget :widget="figure" style="height: 100%; width: 100%" />
     </v-row>
-
   </div>
 </template>
 
 <script>
 </script>
 
-
 <style scoped>
   .only-show-in-tray {
     display: none;
@@ -33,5 +33,4 @@
     width: 100%;
     height: 480px;
   }
-
 </style>

jdaviz/components/toolbar_nested.py

@@ -65,10 +65,13 @@ def __init__(self, viewer, tools_nested, default_tool_priority=[]):
         self.default_tool_priority = default_tool_priority
         self._handle_default_tool()
 
-        for msg in (AddDataMessage, RemoveDataMessage, ViewerAddedMessage, ViewerRemovedMessage,
-                    SpectralMarksChangedMessage):
-            self.viewer.hub.subscribe(self, msg,
-                                      handler=lambda _: self._update_tool_visibilities())
+        # toolbars in the main app viewers need to respond to the data-collection, etc,
+        # but those in plugins do not
+        if hasattr(self.viewer, 'hub'):
+            for msg in (AddDataMessage, RemoveDataMessage, ViewerAddedMessage, ViewerRemovedMessage,
+                        SpectralMarksChangedMessage):
+                self.viewer.hub.subscribe(self, msg,
+                                          handler=lambda _: self._update_tool_visibilities())
 
     def _is_visible(self, tool_id):
         # tools can optionally implement self.is_visible(). If not NotImplementedError

jdaviz/components/toolbar_nested.vue

@@ -4,8 +4,8 @@
         <v-tooltip v-for="[id, {tooltip, img, menu_ind, has_suboptions, primary, visible}] of Object.entries(tools_data)" v-if="primary && visible" bottom>
             <template v-slot:activator="{ on }">
                 <v-btn v-on="on" icon :value="id" style="min-width: 40px !important" @contextmenu="(e) => show_submenu(e, has_suboptions, menu_ind)">
-                    <img :src="img" width="20px" @click.ctrl.stop=""/>
-                    <v-icon small v-if="has_suboptions" class="suboptions-carrot" @click.ctrl.stop="">mdi-menu-down</v-icon>
+                    <img class="invert-if-dark" :src="img" width="20px" @click.ctrl.stop=""/>
+                    <v-icon small v-if="has_suboptions" class="suboptions-carrot invert-if-dark" @click.ctrl.stop="">mdi-menu-down</v-icon>
                 </v-btn>
             </template>
             <span>{{ tooltip }}{{has_suboptions ? " [right-click for alt. tools]" : ""}}</span>
@@ -81,7 +81,7 @@
   }
 </script>
 
-<style>
+<style scoped>
 .suboptions-carrot {
   transform: rotate(-45deg);
   bottom: 0px;

jdaviz/configs/default/plugins/plot_options/plot_options.py

@@ -7,7 +7,6 @@
 from traitlets import Any, Dict, Float, Bool, Int, List, Unicode, observe
 
 from glue.core.subset_group import GroupedSubset
-import bqplot
 from glue.config import stretches
 from glue.viewers.scatter.state import ScatterViewerState
 from glue.viewers.profile.state import ProfileViewerState, ProfileLayerState
@@ -403,8 +402,13 @@ def state_attr_for_line_visible(state):
                                                  'stretch_vmax_value', 'stretch_vmax_sync',
                                                  state_filter=is_image)
 
-        self.stretch_histogram = Plot(self)
-        self.stretch_histogram.add_bins('histogram', color='gray')
+        self.stretch_histogram = Plot(self, viewer_type='histogram')
+        # NOTE: this is a current workaround so the histogram viewer doesn't crash when replacing
+        # data.  Note also that passing x=[0] fails on SOME machines, so we'll pass [0, 1] instead
+        self.stretch_histogram._add_data('ref', x=[0, 1])
+        self.stretch_histogram.layers['ref'].state.visible = False
+        self.stretch_histogram._add_data('histogram', x=[0, 1])
+
         self.stretch_histogram.add_line('vmin', x=[0, 0], y=[0, 1], ynorm=True, color='#c75d2c')
         self.stretch_histogram.add_line('vmax', x=[0, 0], y=[0, 1], ynorm=True, color='#c75d2c')
         self.stretch_histogram.figure.axes[0].label = 'pixel value'
@@ -606,16 +610,6 @@ def _update_stretch_histogram(self, msg={}):
                 y_max = y_limits.max()
 
                 arr = comp.data[y_min:y_max, x_min:x_max]
-
-                size = arr.shape[0] * arr.shape[1]
-                if size > 400**2:
-                    xstep = max(1, round(arr.shape[1] / 400))
-                    ystep = max(1, round(arr.shape[0] / 400))
-                    arr = arr[::ystep, ::xstep]
-                    stretch_hist_downsampled = [size, arr.shape[0] * arr.shape[1]]
-                else:
-                    stretch_hist_downsampled = size
-
                 sub_data = arr.ravel()
 
             else:
@@ -635,49 +629,34 @@ def _update_stretch_histogram(self, msg={}):
 
                 sub_data = comp.data[inds].ravel()
 
-                # downsampling not currently implemented for 2d spectrum
-                stretch_hist_downsampled = len(sub_data)
-
         else:
             # include all data, regardless of zoom limits
             arr = comp.data
-            size = arr.shape[0] * arr.shape[1]
-            if size > 400**2:
-                xstep = max(1, round(arr.shape[1] / 400))
-                ystep = max(1, round(arr.shape[0] / 400))
-                arr = arr[::ystep, ::xstep]
-                stretch_hist_downsampled = [size, arr.shape[0] * arr.shape[1]]
-            else:
-                stretch_hist_downsampled = size
-
             sub_data = arr.ravel()
 
         # filter out nans (or else bqplot will fail)
         if np.any(np.isnan(sub_data)):
             sub_data = sub_data[~np.isnan(sub_data)]
 
-        hist_mark = self.stretch_histogram.marks['histogram']
-        with hist_mark.hold_sync():
-            hist_mark.sample = sub_data
+        self.stretch_histogram._update_data('histogram', x=sub_data)
+
+        if len(sub_data) > 0:
             interval = PercentileInterval(95)
-            if len(sub_data) > 0:
-                hist_lims = interval.get_limits(sub_data)
-                hist_mark.min, hist_mark.max = hist_lims
-                # set the stepsize for vmin/vmax to be approximately 1% of the range of the
-                # histogram (within the percentile interval), rounded to 1-2 significant digits
-                # to avoid random step sizes.  This logic is somewhat arbitrary and can be safely
-                # modified or eventually exposed to the user if that would be useful.
-                stretch_vstep = (hist_lims[1] - hist_lims[0]) / 100.
-                self.stretch_vstep = np.round(stretch_vstep, decimals=-int(np.log10(stretch_vstep))+1)  # noqa
-            hist_mark.bins = self.stretch_hist_nbins
-            # in case only the sample has changed but its length has not,
-            # we'll force the traitlet to trigger a change
-            hist_mark.send_state('sample')
-        if isinstance(stretch_hist_downsampled, list):
-            title = f"{stretch_hist_downsampled[1]} of {stretch_hist_downsampled[0]} pixels"
-        else:
-            title = f"{stretch_hist_downsampled} pixels"
-        self.stretch_histogram.figure.title = title
+            hist_lims = interval.get_limits(sub_data)
+            # set the stepsize for vmin/vmax to be approximately 1% of the range of the
+            # histogram (within the percentile interval), rounded to 1-2 significant digits
+            # to avoid random step sizes.  This logic is somewhat arbitrary and can be safely
+            # modified or eventually exposed to the user if that would be useful.
+            stretch_vstep = (hist_lims[1] - hist_lims[0]) / 100.
+            self.stretch_vstep = np.round(stretch_vstep, decimals=-int(np.log10(stretch_vstep))+1)  # noqa
+
+            self.stretch_histogram.viewer.state.hist_x_min = hist_lims[0]
+            self.stretch_histogram.viewer.state.hist_x_max = hist_lims[1]
+
+        self.stretch_histogram.figure.title = f"{len(sub_data)} pixels"
+
+        # update the n_bins since this may be a new layer
+        self._histogram_nbins_changed()
 
     @observe('is_active', 'stretch_vmin_value', 'stretch_vmax_value', 'layer_selected',
              'stretch_hist_nbins', 'image_contrast_value', 'image_bias_value',
@@ -739,13 +718,7 @@ def _update_stretch_curve(self, msg=None):
                     opacities=[0.5],
                 )
 
-            # reorder marks so histogram is on top:
-            figure_marks = self.stretch_histogram.figure.marks
-            for i, fig_mark in enumerate(figure_marks):
-                if isinstance(fig_mark, bqplot.Bins):
-                    hist_mark = figure_marks.pop(i)
-                    break
-            self.stretch_histogram.figure.marks = figure_marks + [hist_mark]
+            self.stretch_histogram._refresh_marks()
 
     @observe('stretch_vmin_value')
     def _stretch_vmin_changed(self, msg=None):
@@ -756,18 +729,20 @@ def _stretch_vmax_changed(self, msg=None):
         self.stretch_histogram.marks['vmax'].x = [self.stretch_vmax_value, self.stretch_vmax_value]
 
     @observe("stretch_hist_nbins")
-    def _histogram_nbins_changed(self, msg):
-        if self.stretch_histogram is None or msg['new'] == '' or msg['new'] < 1:
+    def _histogram_nbins_changed(self, msg={}):
+        if self.stretch_histogram is None:
             return
-        self.stretch_histogram.marks['histogram'].bins = self.stretch_hist_nbins
-
-    def set_histogram_x_limits(self, x_min=None, x_max=None):
-        # NOTE: leaving this out of user API until API is finalized with interactive setting
-        self.stretch_histogram.set_xlims(x_min, x_max)
+        if self.stretch_hist_nbins == '' or self.stretch_hist_nbins < 1:
+            return
+        self.stretch_histogram.viewer.state.hist_n_bin = self.stretch_hist_nbins
+        # for some reason, this resets the internal marks, so we need to ensure the manual
+        # marks are still plotted
+        self.stretch_histogram._refresh_marks()
 
-    def set_histogram_y_limits(self, y_min, y_max):
+    def set_histogram_limits(self, x_min=None, x_max=None, y_min=None, y_max=None):
         # NOTE: leaving this out of user API until API is finalized with interactive setting
-        self.stretch_histogram.set_ylims(y_min, y_max)
+        self.stretch_histogram.set_lims(x_min=x_min, x_max=x_max,
+                                        y_min=y_min, y_max=y_max)
 
     def _viewer_is_image_viewer(self):
         # Import here to prevent circular import (and not at the top of the method so the import

jdaviz/configs/default/plugins/plot_options/tests/test_plot_options.py

@@ -72,46 +72,47 @@ def test_stretch_histogram(cubeviz_helper, spectrum1d_cube_with_uncerts):
 
     assert po.stretch_histogram is not None
 
-    hist_mark = po.stretch_histogram.marks['histogram']
-    flux_cube_sample = hist_mark.sample
+    hist_lyr = po.stretch_histogram.layers['histogram']
+    flux_cube_sample = hist_lyr.layer.data['x']
 
     # changing viewer should change results
     po.viewer.selected = 'uncert-viewer'
-    assert not allclose(hist_mark.sample, flux_cube_sample)
+    assert not allclose(hist_lyr.layer.data['x'], flux_cube_sample)
 
     po.viewer.selected = 'flux-viewer'
-    assert_allclose(hist_mark.sample, flux_cube_sample)
+    assert_allclose(hist_lyr.layer.data['x'], flux_cube_sample)
 
     # change viewer limits
     fv = cubeviz_helper.app.get_viewer('flux-viewer')
     fv.state.x_max = 0.5 * fv.state.x_max
     # viewer limits should not be affected by default
-    assert_allclose(hist_mark.sample, flux_cube_sample)
+    # (re-retrieve layer - it should not have changed)
+    hist_lyr = po.stretch_histogram.layers['histogram']
+    assert_allclose(hist_lyr.layer.data['x'], flux_cube_sample)
 
     # set to listen to viewer limits, the length of the samples will change
+    # (in this case the layer itself has been replaced)
     po.stretch_hist_zoom_limits = True
-    assert len(hist_mark.sample) != len(flux_cube_sample)
+    hist_lyr = po.stretch_histogram.layers['histogram']
+    assert len(hist_lyr.layer.data['x']) != len(flux_cube_sample)
 
     po.stretch_vmin.value = 0.5
     po.stretch_vmax.value = 1
 
     assert po.stretch_histogram.marks['vmin'].x[0] == po.stretch_vmin.value
     assert po.stretch_histogram.marks['vmax'].x[0] == po.stretch_vmax.value
 
-    assert hist_mark.bins == 25
+    assert po.stretch_histogram.viewer.state.hist_n_bin == 25
     po.stretch_hist_nbins = 20
-    assert hist_mark.bins == 20
+    assert po.stretch_histogram.viewer.state.hist_n_bin == 20
 
-    po.set_histogram_x_limits(x_min=0.25, x_max=2)
-    assert po.stretch_histogram.figure.axes[0].scale.min == 0.25
-    assert po.stretch_histogram.figure.axes[0].scale.max == 2
+    po.set_histogram_limits(x_min=0.25, x_max=2)
+    assert po.stretch_histogram.viewer.state.x_min == 0.25
+    assert po.stretch_histogram.viewer.state.x_max == 2
 
-    po.set_histogram_y_limits(y_min=1, y_max=2)
-    assert po.stretch_histogram.figure.axes[1].scale.min == 1
-    assert po.stretch_histogram.figure.axes[1].scale.max == 2
-
-    po.stretch_vmin.value = hist_mark.min - 1
-    po.stretch_vmax.value = hist_mark.max + 1
+    po.set_histogram_limits(y_min=1, y_max=2)
+    assert po.stretch_histogram.viewer.state.y_min == 1
+    assert po.stretch_histogram.viewer.state.y_max == 2
 
     assert po.stretch_histogram.marks['vmin'].x[0] == po.stretch_vmin.value
     assert po.stretch_histogram.marks['vmax'].x[0] == po.stretch_vmax.value