Statistics and key frequency

Two panels summarize a whole document: Statistics counts its nodes by type and depth, and Key frequency counts its keys. Open them from the activity rail on the right of the window. Both update when the document changes, for example after a reload or an edit.

Statistics panel

The Statistics panel shows:

• nodes: the number of objects, arrays, and values in the document.
• max depth: how many levels deep the document goes.
• Type Distribution: the number of objects, arrays, strings, numbers, booleans, and nulls, as bars.
• Top Keys: the 20 most frequent key names, with how often each occurs.
• Depth Distribution: the number of nodes at each level, from d=0 for the root.

Key frequency panel

The Key frequency panel lists the key names of the document, most frequent first, with a bar and a count for each. It counts every key at every depth, by name: id in a user and id in an order count together. The list shows the 100 most frequent keys.

To find where a key occurs, search for it with Keys and Exact. To count the values of a field across the rows of an array, use Value frequency in the Table view.

On a very large file, the panel takes a while: it reads every key in the file.

Approximate results on large files

The Statistics panel does not always count every node. In a JSON document with very large objects or arrays, it counts the part JSON Digger has indexed, which leaves out deep or capped children; in an NDJSON file, it reads a sample of the records. It then shows indexed nodes instead of nodes, and the note “Based on the indexed sample”. The counts and bars show the shape of the document, not exact totals. Key frequency always counts every key. For the NDJSON sample size, see Limits.