Hi! I’ve been diving into Metabase a bit this week and I’m a huge fan so far! Really impressed with its features. I’m part of the City of Philadelphia’s Open Data program, and we help departments publish government data - like crime incidents, employee salaries, police complaints, etc. - to increase government transparency.
In addition to publishing the raw (CSV) data and an API, we try to accompany our data releases with a visualisation to make them more accessible to less technical users. We use a tool built in-house called VizWit. It primarily shows 3-5 charts grouped by key categories (e.g. crime category, date, zip code) and the raw data in a table, but the most significant feature it offers is cross-filtering: clicking on one of the categories will filter the other charts to only show the data within that category. Here’s an illustration:
This feature enables users to “drill down” on the data and answer questions such as, “How many thefts were there in my neighborhood last month?” without having to know what to do with a giant CSV file. As a result, our hope is that providing these visualisations broadens the audience for open data.
We would love to use Metabase (public dashboards) instead because of all the additional power it offers, plus the fact that we don’t have a lot of bandwidth to iterate on our in-house tool. But I can’t seem to accomplish the same cross-filtering functionality with Metabase. The built-in filtering feature seems to work quite differently, and requires you search for categories.
Is there any interest in cross-filtering for Metabase? If so, is there a way to collaborate on a feature like this?