Query and Transform

A quick tour of the SDK and catalog server

Joins, filters, and aggregations on Rerun data

Query and decode image frames from a catalog

Filter by episode, time, and column to focus a dataframe

Query video streams & decode video frames

Fill forward to run compute at a fixed frequency

Generate clickable URLs that open segments in the Viewer

Load Rerun data into Pandas, Polars, or DuckDB

Resample data streams produced at different frequencies

Tips for keeping queries fast as datasets grow

Reference a subset of an existing dataset without copying

Persist derived data from catalog dataset queries