Vizro looks nice, as does Evidence (mentioned below). I think it's good to remember that these types of apps often don't exist in a silo. It would be excellent to be able to integrate them into a larger platform (Django app, etc) to create a less disjointed experience for consumers.
Some discussions on other McKinsey contributions: https://news.ycombinator.com/item?id=33869800 https://news.ycombinator.com/item?id=31975877 https://news.ycombinator.com/item?id=28746538
This look pretty good !
I'm a heavy user of data vis, and this looks worth trying.
Looks like McKinsey is doing and open sourcing some interesting things - I've also tried Kedro and while it feels a bit 'heavy' to use, I like a lot the emphasis on structuring the data and steps.
I am happy to see more libraries in this space. I do not like R, but the Shiny workflow is so much faster and easier for novices to bang out some value. Streamlit is really heavy in comparison.
Minor annoyance: new project and it chose YAML over TOML.
This is interesting. I've built reports in Dash before. As I remember it was kind of tricky to hook up the filters etc, and "callbacks" confused me for a while. Is that the kind of thing Vizro addresses?
Do you have a "why Vizro" kind of explanation somewhere? Would be really helpful early in the docs, or on the GH readme.
So, like streamlit?
I don't get it. Why not just use Plotly and Dash directly, neither of which are especially hard? This seems like a pointless layer of abstraction.
If you are interested in this, but would prefer to define reports with a markup language (and SQL), I work on an open source code-based BI tool called Evidence, which might be of interest to you.
It's effectively a static site generator aimed at building automated reports and analysis.
https://github.com/evidence-dev/evidence
Previous discussions on HN:
https://news.ycombinator.com/item?id=28304781 - 91 comments
https://news.ycombinator.com/item?id=35645464 - 97 comments