How to shade an area of a catagorical bar chart?

[RobM]

Hello,
I would like to shade an area of a categorical bar chart but get the error:
AttributeError: ‘Figure’ object has no attribute ‘varea’
Does this mean it is not possible to use vbar and varea together?
This is a minimal example:

from bokeh.io import output_file, show
from bokeh.plotting import figure

output_file("bars.html")

fruits = ['Apples', 'Pears', 'Nectarines', 'Plums', 'Grapes', 'Strawberries']
counts = [5, 3, 4, 2, 4, 6]

p = figure(x_range=fruits, plot_height=250, title="Fruit counts",
           toolbar_location=None, tools="")

p.vbar(x=fruits, top=counts, width=0.9)
p.varea(x=fruits, y1=2, y2=3)
p.xgrid.grid_line_color = None
p.y_range.start = 0

show(p)

[_jm]

@RobM

Your minimal example works as expected (tested with bokeh 2.2.3 and Chrome 89.0.4389.114).

Here’s the output with the code as provided …

And here’s the output with the p.varea statement commented-out.

[RobM]

Ah, ok thanks for checking. It looks like I’m stuck on verison 1.0.4 which I guess is the problem.

[Bryan]

varea, harea, etc were added in version 1.2 so you would need to update to at least that.

[RobM]

Hi,
Thanks I’ve updated my environment to one using a more recent version of Bokeh. I’d like to be able to use a hover tool to inspect the area but can’t work out how to do this. Would I need to add a second hover tool some how?

rom bokeh.io import output_file, show
from bokeh.plotting import figure
from bokeh.models.tools import HoverTool

output_file("bars.html")

fruits = ['Apples', 'Pears', 'Nectarines', 'Plums', 'Grapes', 'Strawberries']
counts = [5, 3, 4, 2, 4, 6]
tooltips=[('Index', '$index'), ('Count', '@top')]
p = figure(x_range=fruits, plot_height=250, title="Fruit counts", tools=['hover'], tooltips=tooltips)
hover_tool = p.select(type=HoverTool) 
hover_tool.names = ["bar"]
p.vbar(x=fruits, top=counts, width=0.9, name='bar')
p.varea(x=[0, len(fruits)], y1=2, y2=3, color='red', alpha=0.4, name='area')
p.xgrid.grid_line_color = None
p.y_range.start = 0

show(p)

[Bryan]

@RobM You need to describe what you want to accomplish in more detail, I don’t understand your question.

[RobM]

Sorry, perhaps it isn’t possible with varea but I have an example sort of working with patches. What I would like to know is can I specify separate tooltips for the different glyphs? I’d like to avoid the green patch hover tool box showing “Count:???”.

from bokeh.io import output_file, show
from bokeh.plotting import figure
from bokeh.models.tools import HoverTool

output_file("bars.html")

fruits = ['Apples', 'Pears', 'Nectarines', 'Plums', 'Grapes', 'Strawberries']
counts = [5, 3, 4, 2, 4, 6]
tooltips=[('Index', '$index'), ('Count', '@top')]

p = figure(x_range=fruits, plot_height=250, title="Fruit counts", tools=['hover'], tooltips=tooltips)
hover_tool = p.select(type=HoverTool) 
hover_tool.names = ["bar", "patch"]

p.vbar(x=fruits, top=counts, width=0.9, name='bar')
p.varea(x=[0, len(fruits)], y1=2, y2=3, color='red', alpha=0.4, name='area')
p.patches([[0,0,len(fruits),len(fruits)]],[[3,4,4,3]], color='green', alpha=0.4, name='patch')
p.xgrid.grid_line_color = None
p.y_range.start = 0

show(p)

[Bryan]

By default, hover tools operate on all available renderers. But you can add multiple different hover tools, and restrict each to only a specific glyph (or glyphs) by configring their renderers property

[RobM]

thanks that was what I was hoping, but I haven’t worked out how to do it yet. Tools and tooltips are added when setting up the figure, how do I go about adding a second hover tool?

[_jm]

@RobM

@RobM

Tools and tooltips are added when setting up the figure, how do I go about adding a second hover tool?

That’s certainly a common way to include a hover tool but not the only way.

You can create a hover tool and specify its properties including the renderers property as a list that you want to operate on / restrict focus to.

So something like the following where r0 and r1 are two renderers previously created each assigned to its own HoverTool h0 and h1, respectively.

The last line in this sketch/outline assigns both tools to a figure p.

from bokeh.models import HoverTool
...
h0 = HoverTool(renderers=[r0], ...)
h1 = HoverTool(renderers=[r1], ...)
_ = [p.add_tools(tool) for tool in (h0, h1)]

[RobM]

Thanks, I’ll give that a go. I had been trying things like this

p = figure(x_range=fruits, plot_height=250, title="Fruit counts", tools=['hover, hover'])

hover0 = p.select(type=HoverTool) 
hover0.names = ["bar"]
hover0.tooltips = [('Index', '$index'), ('Count', '@top')]

hover1 = p.select(type=HoverTool) 
hover1.names = ["patch"]
hover1.tooltips = [('Index', '$index'), ('Y2', '6')]

[Bryan]

The tools (or hover) args to figure are conveniences for the common case. They only add a single HoverTool to the plot, so then p.select(type=HoverTool) will return the same hover tool instance every time you call it. If you need multiple hover tools then they have to be added manually / individually as @_jm demonstrated. More documentation is here

Configuring plot tools — Bokeh 2.4.2 Documentation

[RobM]

Thanks all, I have a working example now that I think I can work from.

from bokeh.io import output_file, show
from bokeh.plotting import figure
from bokeh.models.tools import HoverTool

output_file("bars.html")

fruits = ['Apples', 'Pears', 'Nectarines', 'Plums', 'Grapes', 'Strawberries']
counts = [5, 3, 4, 2, 4, 6]

p = figure(x_range=fruits, plot_height=250, title="Fruit counts", tools="")

bar = p.vbar(x=fruits, top=counts, width=0.9)
patch = p.patches([[0,0,len(fruits),len(fruits)]],[[3,4,4,3]], color='green', alpha=0.4)

p.xgrid.grid_line_color = None
p.y_range.start = 0

h0 = HoverTool(renderers=[bar]) 
h0.tooltips = [('Index', '$index'), ('Count', '@top')]
p.add_tools(h0)

h1 = HoverTool(renderers=[patch]) 
h1.tooltips = [('Count', '3 to 4')]
p.add_tools(h1)

show(p)

[RobM]

although I’m finding patches a lot tricker to use than varea with a columnDataSource