This is a package I’ve been working on this summer to make dashboarding easier. It combines Plotly Dash with Jinja templating, and attempts to add a convenient abstraction layer on top of the Dash components. I’d love to hear what a few inquisitive minds think about it.
See it here at the Bento repo on Github
Also, I just recently put together the beginner tutorial within the docs, if anyone is willing to be a guinea pig.
This is very cool, thanks for sharing! Would love to hear more about the architecture. Is the idea that the dashboard widgets are autogenerated from the dataset specification?
Yes, if your dataframe is a simple case and the code is working correctly, some of the widgets will be entirely hands-free. Categorical columns would be detected and the unique values present in them would automatically populate dropdowns. For example, this is true for the sample covid dashboard I made with Bento, for the “selector” widget (e.g. where you “Select country”, etc).
Broadly, though, I’ve tried to incorporate many different things that I had to learn to get some of the functionality I wanted: easy multiple page support, styling the components based on a theme, component loading overlays, the grid system, etc. So it’s maybe more about a frictionless experience than any one individual feature. Easier said than done though.
I do hope it proves useful!