Turning off Tick Marks in Bokeh

Question

I am working on the Bokeh (0.6.1) tutorial and trying to turn off the tick marks and labels in one of the exercise plots, the scatter plot:

from __future__ import division

import numpy as np
from six.moves import zip
from bokeh.plotting import *
from bokeh.objects import Range1d

output_file("scatter.html")

figure()

N = 4000

x = np.random.random(size=N) * 100
y = np.random.random(size=N) * 100
radii = np.random.random(size=N) * 1.5
colors = [
    "#%02x%02x%02x" % (r, g, 150) 
    for r, g in zip(np.floor(50+2*x), np.floor(30+2*y))
]

circle(x, y, radius=radii,
       fill_color=colors, fill_alpha=0.6,
       line_color=None, Title="Colorful Scatter")

grid().grid_line_color = None
axis().axis_line_color = None

# QUESTION PART 1: Is this the right way to turn off tick labels?
axis().major_label_text_font_size = '0pt'  
# QUESTION PART 2: ...and how to turn off tick marks also?

show()  # open a browser

I have managed to turn off the tick labels but no amount of searching the documentation and googling has revealed the incantation needed to turn off the tick marks.

Also I am not sure that setting axis().major_label_text_font_size to 0pt is the right way to turn off tick labels or if it is a kludge. Nothing else seemed to work.

Am I missing something obvious?

Steven C. Howell · Accepted Answer · 2017-02-06T15:40:02.093

This answer is an update for the more recent 0.12.4 version of Bokeh. For additional information, these commands are taken from the Styling Visual Attributes page of the Bokeh documentation.

To turn off the major and minor tick marks set their color to None:

p = bokeh.plotting.figure(plot_width=400, plot_height=400)
p.circle([1,2,3,4,5], [2,5,8,2,7], size=10)

p.xaxis.major_tick_line_color = None  # turn off x-axis major ticks
p.xaxis.minor_tick_line_color = None  # turn off x-axis minor ticks

p.yaxis.major_tick_line_color = None  # turn off y-axis major ticks
p.yaxis.minor_tick_line_color = None  # turn off y-axis minor ticks

To turn off the tick labels set the font size to '0pt':

p.xaxis.major_label_text_font_size = '0pt'  # turn off x-axis tick labels
p.yaxis.major_label_text_font_size = '0pt'  # turn off y-axis tick labels

A similar result can be achieved by setting the font color to `None', with the disadvantage that space is still maintained for the tick labels.

p.xaxis.major_label_text_color = None  # turn off x-axis tick labels leaving space
p.yaxis.major_label_text_color = None  # turn off y-axis tick labels leaving space

This code snippet exemplifies removing both the major and minor tick lines and also the tick labels.

import bokeh.io
import bokeh.plotting
import bokeh.layouts
bokeh.io.output_file('remove_tick_marks.html')

p0 = bokeh.plotting.figure(plot_width=200, plot_height=200, 
                           x_axis_label='x', y_axis_label='y', 
                           title='original')
p0.circle([1,2,3,4,5], [2,5,8,2,7], size=10)

p1 = bokeh.plotting.figure(plot_width=200, plot_height=200, 
                           x_axis_label='x', y_axis_label='y', 
                           title='remove tick marks')
p1.circle([1,2,3,4,5], [2,5,8,2,7], size=10)
p1.xaxis.major_tick_line_color = None  # turn off x-axis major ticks
p1.xaxis.minor_tick_line_color = None  # turn off x-axis minor ticks
p1.yaxis.major_tick_line_color = None  # turn off y-axis major ticks
p1.yaxis.minor_tick_line_color = None  # turn off y-axis minor ticks

p2 = bokeh.plotting.figure(plot_width=200, plot_height=200, 
                           x_axis_label='x', y_axis_label='y', 
                           title='remove tick labels')
p2.circle([1,2,3,4,5], [2,5,8,2,7], size=10)
p2.xaxis.major_tick_line_color = None  # turn off x-axis major ticks
p2.xaxis.minor_tick_line_color = None  # turn off x-axis minor ticks
p2.yaxis.major_tick_line_color = None  # turn off y-axis major ticks
p2.yaxis.minor_tick_line_color = None  # turn off y-axis minor ticks
p2.xaxis.major_label_text_font_size = '0pt'  # preferred method for removing tick labels
p2.yaxis.major_label_text_font_size = '0pt'  # preferred method for removing tick labels

p3 = bokeh.plotting.figure(plot_width=200, plot_height=200, 
                           x_axis_label='x', y_axis_label='y', 
                           title='notice extra space')
p3.circle([1,2,3,4,5], [2,5,8,2,7], size=10)
p3.xaxis.major_tick_line_color = None  # turn off x-axis major ticks
p3.xaxis.minor_tick_line_color = None  # turn off x-axis minor ticks
p3.yaxis.major_tick_line_color = None  # turn off y-axis major ticks
p3.yaxis.minor_tick_line_color = None  # turn off y-axis minor ticks
p3.xaxis.major_label_text_color = None  #note that this leaves space between the axis and the axis label  
p3.yaxis.major_label_text_color = None  #note that this leaves space between the axis and the axis label  

grid = bokeh.layouts.gridplot([[p0, p1, p2, p3]])
bokeh.io.show(grid)

Did not quite work for me. When I converted SVG to PDF, Adobe Acrobat complained. After investigations, turns out the ticks and labels are not gone, and Adobe Acrobat renders them the way it wants (perhaps, it has some minimum font size). — Dmytro Bogatov, Feb 14 '20 at 00:32

daedalus · Answer 2 · 2014-12-04T03:59:35.843

6

I am not sure if the absence of an answer over more than a week is due to people not knowing it, or because the question is being ignored as too obvious.

Anyway, in the hope that others might find it useful, I post this answer. I have found a way of doing it that seems so much like a hack that I am only posting it in the hope that someone will improve on it...

from __future__ import division

import numpy as np
from six.moves import zip
from bokeh.plotting import *

output_file("scatter.html")

figure()

N = 4000

x = np.random.random(size=N) * 100
y = np.random.random(size=N) * 100
radii = np.random.random(size=N) * 1.5
colors = ["#%02x%02x%02x" % (r, g, 150) 
          for r, g in zip(np.floor(50+2*x), np.floor(30+2*y))]

circle(x, y, radius=radii,
       fill_color=colors, fill_alpha=0.6,
       line_color=None, Title="Colorful Scatter")

grid().grid_line_color = None
axis().axis_line_color = None
curplot().outline_line_color = None

# Turn off tick labels
axis().major_label_text_font_size = '0pt'  
# Turn off tick marks 
axis().major_tick_line_color = None  # turn off major ticks
axis()[0].ticker.num_minor_ticks = 0  # turn off minor ticks
axis()[1].ticker.num_minor_ticks = 0

show()  # open a browser

edited Dec 04 '14 at 03:59

answered Dec 04 '14 at 03:43

daedalus

10,873
5
50
71

No, this is probably the current best way to do it. We could probably add another keyword arg to `figure` but its a balancing act, adding convenience but avoiding "parameter soup". If you have suggestions for improving the spelling, etc. I encourage you to submit them on the Bokeh GH issue tracker, such input is always welcome. – bigreddot Dec 25 '14 at 00:37
Athough, you should also be able to set `major_label_text_font_color = None` and have that work as well for removing major labels. I think that is my preference over setting the font size but I will concede there is not much difference. – bigreddot Dec 25 '14 at 00:40
Thanks @bigreddot, it is reassuring to see I am not missing the obvious. I come from heavy use of ggplot2 and expected to find a simple switch as in http://www.cookbook-r.com/Graphs/Axes_(ggplot2)/ but I am satisfied with this interim solution and I will consider filing an issue if I have a helpful suggestion to make. – daedalus Dec 25 '14 at 18:18
@bigreddot: I think the property is called `major_label_text_color` (without the `font`). In my eyes, gauden's approach for removing the labels has the advantage that labels with a font size of 0pt don't take up room anymore, whereas the labels that have their color set to `None` still affect the margin width. – Jake Oct 30 '15 at 08:39
is it possible to turn of major labour text color lables for just one axis and not both – Francisco Vargas Feb 19 '16 at 01:37
Not sure this is possible... Do consider asking this as a separate question. – daedalus Feb 20 '16 at 13:23

Turning off Tick Marks in Bokeh

2 Answers2