CODEWITHSUNDEEP
githubipythonjupyterjupyter-notebookbokeh
ShanZhengYang
ShanZhengYang

Reputation: 17631

Why doesn't BokehJS show in a rendered Jupyter/IPython notebook on GitHub?

I have the following code in a Jupyter notebook:

import numpy as np
from bokeh.plotting import figure, show
from bokeh.io import output_notebook
N = 4000
x = np.random.random(size=N) * 100
x = 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))]
output_notebook()
# Loading BokehJS ...
p = figure()
p.circle(x, y, radius=radii, fill_color=colors, fill_alpha=0.6, line_color=None)
show(p)

However, it does not show any plot or graphics; it simply is stuck on "Loading BokehJS".

In principle, this should work with nbviewer, as rendered notebooks are stripped of all Javascript on GitHub (I think?). In my experience however, it doesn't.

Upvotes: 1

Views: 654

Answers (1)

bigreddot
bigreddot

Reputation: 34568

GitHub scrubs all JavaScript from all Jupyter notebooks before rendering them (presumably for security reasons). Bokeh requires JavaScript code from the client library BokehJS in order to render or do anything at all. Given this, I would not expect Bokeh plots in Jupyter notebooks to ever work on GitHub, unfortunately.

I would very much like for it to be workable, but it is entirely outside our control. I have reached out to GitHub asking for an option to disable rendering entirely for notebooks in a repo, on the reasoning that "not rendering at all" is preferable to "rendering but looking broken" but have not heard back from them.

Note that nbviewer does not strip JavaScript, which is why all the notebooks at the Bokeh nbviewer.org gallery show up just fine.

Upvotes: 3

Related Questions