X-raying your data

X-rays are Metabase's way of guessing what's important about your data and creating a dashboard of questions for you. They take advantage of the fact that most questions involve either counting rows or summing a numeric column, grouping by a date column, and potentially grouping by another low cardinality column. The actual feature is more complex than that, involving some machine learning behind the scenes, but that's generally how they work.

While I don't believe that x-rays are an adequate replacement for a curious data scientist, I do think they're useful when used by a curious data scientist. Because x-rays are simple to create and fast to materialize, they make for a good base layer of simple questions that you can use or build upon to make more complex questions. Let's see an example.

Creating an x-ray

To get started, click Browse Data | Pies. Note that as you hover over each table in the pies database, an icon of a...