Nested categories, source = ColumnDataSource problem with Bokeh: - nested

Question

I just can't see what I'm doing wrong. Please help me.
I would like to output a bar chart for nested categories.
If I comment out source, and I don't add source=source in p1.vbar, I get something... but I can't figure out for the life of me how to get the bars to render.
Thank you for your help and time.
Here is the code:
colors=['#90be6d','#adb5bd','#d90429']
x = [('positive', 'k1'), ('positive', 'k2'),
('positive', 'k3'), ('positive', 'k4'),
('positive', 'k5'), ('positive', 'k6'),
('positive', 'k7'), ('positive', 'k8'),
('positive', 'k9'), ('positive', 'k10'),
('positive', 'k11'), ('positive', 'k12'),
('positive', 'k13'),
('neutral', 'k1'), ('neutral', 'k2'),
('neutral', 'k3'), ('neutral', 'k4'),
('neutral', 'k5'), ('neutral', 'k6'),
('neutral', 'k7'), ('neutral', 'k8'),
('negative', 'k1'), ('negative', 'k2'),
('negative', 'k3'), ('negative', 'k4'),
('negative', 'k5'), ('negative', 'k6'),
('negative', 'k7'), ('negative', 'k8'),
('negative', 'k9'), ('negative', 'k10'),
('negative', 'k11')]
counts = (404, 190, 174, 213, 178, 152, 146, 173, 97,
81, 88, 77, 144, 60, 26, 31, 30, 44, 42, 32,
135, 302, 88, 68, 96, 72, 87, 72, 47, 59, 42, 113)
source = ColumnDataSource(data=dict(x=x, counts=counts)) # here is my problem...
p1 = figure(x_range=FactorRange(*x), plot_height=400, title="test",
toolbar_location=None, tools="")
p1.vbar(x='x', top='counts', width=.9, color=colors, source=source)
p1.y_range.start = 0
p1.x_range.range_padding = 0.1
p1.xaxis.major_label_orientation = 1
p1.xgrid.grid_line_color = None
show(p1)
Thank you.

Answer 1

Your problem is not with the source but with colors. When you run your code, you should see this error:
Traceback (most recent call last):
File "/.../test.py", line 33, in <module>
p1.vbar(x='x', top='counts', width=.9, color=colors, source=source)
File "/.../bokeh/plotting/_decorators.py", line 54, in wrapped
return create_renderer(glyphclass, self, **kwargs)
File "/.../bokeh/plotting/_renderer.py", line 94, in create_renderer
raise RuntimeError(_GLYPH_SOURCE_MSG % nice_join(incompatible_literal_spec_values, conjuction="and"))
RuntimeError:
Expected fill_color and line_color to reference fields in the supplied data source.
When a 'source' argument is passed to a glyph method, values that are sequences
(like lists or arrays) must come from references to data columns in the source.
For instance, as an example:
source = ColumnDataSource(data=dict(x=a_list, y=an_array))
p.circle(x='x', y='y', source=source, ...) # pass column names and a source
Alternatively, *all* data sequences may be provided as literals as long as a
source is *not* provided:
p.circle(x=a_list, y=an_array, ...) # pass actual sequences and no source
It's pretty self-descriptive, except for one part - it mentions fill_color and line_color argument which you don't use. The thing is, specifying color for vbar is actually a shorthand for specifying both fill_color and line_color, that's why the error mentions them.
What you gotta do to fix the issue is to provide the colors in a way that Bokeh understands. Either create a separate color column in the data source (it has to be the same length as all other columns, so you'll have to repeat some values) or create a categorical color mapper. An example of the latter can be seen here: https://docs.bokeh.org/en/latest/docs/user_guide/categorical.html?highlight=categoricalcolormapper#colors