Improve documentation for creating interactive plots in Notebook
See original GitHub issueI’m trying to figure out how to create a fully interactive plots in Jupyter Notebook, with widgets, etc. I’ve spent several days going through the Bokeh documentation, trying to wrap my head around the paradigm. There seems to be three ways of doing this:
- JS callbacks
- Bokeh server
- Embed Bokeh application in Notebook
I’m wondering why there’s not a more straight-forward way of doing this, though. The plots created with output_notebook
are interactive, so that means there’s already an event loop there. Why can’t that be used for widget callbacks written in Python?
No so much a feature request as trying to fill in what appear to be gaps in the documentation. Maybe the gaps are for historical reasons, e.g. Bokeh evolved with Notebook support as an afterthought. I’m willing to help improve the Bokeh documentation for Notebook, if I can get clearer idea of how this all works, and why you can’t simply add widgets and callbacks in Notebook.
There doesn’t seem to be a fully featured Python plotting library that integrates well with Notebook, but Bokeh is the best candidate I’ve found. But, at present, it takes quite a bit of time in the documentation just to get started.
Issue Analytics
- State:
- Created 5 years ago
- Comments:13 (12 by maintainers)
Top GitHub Comments
@okomarov I only became aware of it after reading this blog post. The author points out that this feature is only available as of release
0.12.5
. I think this information should be added to the docs, after confirming that it’s correct.Posts on SO and blogs are great, but I wish more people would contribute to the actual docs.
There are a few answers on SO that give examples of bokeh server within notebook. Here’s one I made sometime ago. Unfortunately I guess they tend to pop-up on specific use cases search rather than looking for notebook/server answers.