After studying Panel and Bokeh for a couple of weeks I thought I would try out using Bokeh in Streamlit to create som Bokeh plots via st.bokeh_chart.
What I found surprised me. It turns out you can in fact use Bokeh plotting, layouts, some widgets like tables and tabs and JS Calbacks!.
And even more thought provoking. I think the Bokeh server can integrate well with Streamlit to get advanced Python Callbacks as well. And even more thought provoking. I think we could actually get Panel integrated as well! I just haven’t figured out how to do it yet
Whether it’s a good thing I’ve discovered or it’s Pandoras Box I’ve opened I don’t yet know. What do you think?
Please could you provide source code for this application? Or teach me how you do that, was amazing this tabs feature, wish it exists in streamlit itself and i wish to use that tab application.
i tried awesome-streamlit.org but doesn’t open here, i think the website has a problem.
@Marc panel integration with streamlit would be extremely useful, the most relevant example is in displaying large raster datasets. Holoviz/Datashader (with panel for live interactions) are best in class for this use case but it really requires to the client/server interactivity for it to work in a meaningful way.
In the meantime using bokeh to expand the layout capabilities of streamlit is genius