add option to plot probabilities with stacked bars

Author: bjarthur

configuration.py

@@ -68,6 +68,7 @@
 gui_spectrogram_low_hz=0
 gui_spectrogram_high_hz=1250
 gui_spectrogram_clip=[1,99]
+gui_probability_style="lines" # either "lines" or "bars"
 
 # neural network architecture to use
 architecture_plugin="convolutional"

src/gui/model.py

@@ -248,6 +248,7 @@ def init(_bokeh_document, _configuration_file, _use_aitch):
     global time_scale, freq_scale, context_time_scale, context_freq_scale
     global context_waveform_low, context_waveform_high, label_colors
     global spectrogram_colormap, spectrogram_clip, spectrogram_window, spectrogram_length_sec, spectrogram_overlap, spectrogram_low_hz, spectrogram_high_hz
+    global probability_style
     global overlapped_prefix
     global deterministic
     global context_width_sec0, context_offset_sec0
@@ -376,6 +377,8 @@ def is_local_server_or_cluster(varname, varvalue):
     context_freq_units = gui_context_freq_units
     context_freq_scale = gui_context_freq_scale
 
+    probability_style=gui_probability_style
+
     spectrogram_clip=gui_spectrogram_clip
     spectrogram_colormap=gui_spectrogram_colormap
     spectrogram_window=gui_spectrogram_window

src/gui/view.py

@@ -2,9 +2,9 @@
 import sys
 from bokeh.models.widgets import RadioButtonGroup, TextInput, Button, Div, DateFormatter, TextAreaInput, Select, NumberFormatter, Slider, Toggle, ColorPicker, MultiSelect, Paragraph
 from bokeh.models.formatters import FuncTickFormatter
-from bokeh.models import ColumnDataSource, TableColumn, DataTable, LayoutDOM, Span
+from bokeh.models import ColumnDataSource, TableColumn, DataTable, LayoutDOM, Span, HoverTool
 from bokeh.plotting import figure
-from bokeh.transform import linear_cmap
+from bokeh.transform import linear_cmap, stack
 from bokeh.events import Tap, DoubleTap, PanStart, Pan, PanEnd, ButtonClick, MouseWheel
 from bokeh.models.callbacks import CustomJS
 from bokeh.models.markers import Circle
@@ -1130,9 +1130,19 @@ def context_update():
         if xwav and not np.isnan(xwav[0][-1]):
             spectrogram_range_source.data.update(x=[xwav[0][-1]])
 
-    probability_source.data.update(xs=xprob, ys=yprob,
-                                   colors=[M.label_colors[x] for x in M.used_labels],
-                                   labels=list(M.used_labels))
+    if M.probability_style=="lines":
+        probability_source.data.update(xs=xprob, ys=yprob,
+                                       colors=[M.label_colors[x] for x in M.used_labels],
+                                       labels=list(M.used_labels))
+    else:
+        if len(xprob)>1:
+            probability_source.data.update(**{'x'+str(i):(xprob[i] if i<len(xprob) else xprob[0])
+                                              for i in range(M.nlabels)},
+                                           **{'y'+str(i):(yprob[i] if i<len(yprob) else [0]*len(yprob[0]))
+                                              for i in range(M.nlabels)})
+            if len(xprob[0])>1:
+                for g in probability_glyphs:
+                    g.glyph.width = xprob[0][1] - xprob[0][0]
 
     if M.context_waveform:
         waveform_quad_grey_used.data.update(left=left_used,
@@ -1248,6 +1258,9 @@ def recordings_update():
         recordings.options = []
 
     M.used_labels = set([x['label'] for x in M.used_sounds]) if M.used_sounds else []
+    if M.probability_style=="bars":
+        for ilabel,label in enumerate(M.used_labels):
+            probability_glyphs[ilabel].name = label
     M.label_colors = { l:c for l,c in zip(M.used_labels, cycle(label_palette)) }
     M.isnippet = -1
     M.context_sound = None
@@ -1519,7 +1532,7 @@ def init(_bokeh_document):
     global p_snippets, snippet_palette, snippets_dy, snippets_both, snippets_label_sources_clustered, snippets_label_sources_annotated, snippets_wave_sources, snippets_wave_glyphs, snippets_gram_sources, snippets_gram_glyphs, snippets_quad_grey, snippets_quad_fuchsia
     global p_waveform, waveform_span_red, waveform_quad_grey_used, waveform_quad_grey_annotated, waveform_quad_grey_pan, waveform_quad_fuchsia, waveform_source, waveform_glyph, waveform_label_source_used, waveform_label_source_annotated
     global p_spectrogram, spectrogram_span_red, spectrogram_quad_grey_used, spectrogram_quad_grey_annotated, spectrogram_quad_grey_pan, spectrogram_quad_fuchsia, spectrogram_source, spectrogram_glyph, spectrogram_label_source_used, spectrogram_label_source_annotated, spectrogram_range_source, spectrogram_length
-    global p_probability, probability_span_red, probability_source, probability_glyph
+    global p_probability, probability_span_red, probability_source
     global which_layer, which_species, which_word, which_nohyphen, which_kind
     global color_picker
     global zoom_width, zoom_offset, zoomin, zoomout, reset, panleft, panright, allleft, allout, allright, firstlabel, nextlabel, prevlabel, lastlabel
@@ -1750,9 +1763,12 @@ def init(_bokeh_document):
                    text_font_size='6pt', text_align='center', text_baseline='bottom',
                    text_line_height=0.8, level='underlay', text_color='white')
 
-    TOOLTIPS = """
-        <div><div><span style="color:@colors;">@labels</span></div></div>
-    """
+    if M.probability_style=="lines":
+        TOOLTIPS = """
+            <div><div><span style="color:@colors;">@labels,$y</span></div></div>
+        """
+    else:
+        TOOLTIPS = ""
 
     p_probability = figure(plot_width=M.gui_width_pix, tooltips=TOOLTIPS,
                            plot_height=M.context_probability_height_pix,
@@ -1765,9 +1781,27 @@ def init(_bokeh_document):
     p_probability.y_range.end = 1
     p_probability.xaxis.visible = False
 
-    probability_source = ColumnDataSource(data=dict(xs=[], ys=[], colors=[], labels=[]))
-    probability_glyph = p_probability.multi_line(xs='xs', ys='ys',
-                                                 source=probability_source, color='colors')
+    if M.probability_style=="lines":
+        probability_source = ColumnDataSource(data=dict(xs=[], ys=[], colors=[], labels=[]))
+        global probability_glyph 
+        probability_glyph = p_probability.multi_line(xs='xs', ys='ys',
+                                                     source=probability_source, color='colors')
+    else:
+        xs = {'x'+str(i):[] for i in range(M.nlabels)}
+        ys = {'y'+str(i):[] for i in range(M.nlabels)}
+        probability_source = ColumnDataSource(data=xs|ys)
+        global probability_glyphs
+        probability_glyphs = []
+        for i in range(M.nlabels):
+            probability_glyphs.append(p_probability.vbar(
+                    x='x'+str(i),
+                    bottom=stack(*['y'+str(j) for j in range(i)]),
+                    top=stack(*['y'+str(j) for j in range(i+1)]),
+                    line_width = 0,
+                    fill_color=label_palette[i],
+                    source=probability_source))
+            p_probability.add_tools(HoverTool(renderers=[probability_glyphs[i]],
+                                              tooltips=[("", "$name, @y"+str(i))]))
 
     probability_span_red = Span(location=0, dimension='height', line_color='red')
     p_probability.add_layout(probability_span_red)