How-to guides for sending data to Rerun.
Send entire columns at once
Log many rows of data in a single call
Send dataframes to a recording
Log a whole table into a recording in a single call
Using layers to append data to segments
Append data to existing segments via named layers
Send partial updates over time
Update only the components that change between frames
Send tables to Rerun
Send tables as dataframes to the Rerun Viewer
Share recordings across multiple processes
Log to a single recording from multiple processes, threads, machines, or languages
Log to several recordings from one process
Create more than one recording from a single process, and why calling rr.init again replaces the global recording
Send user-defined data
Log custom data via AnyValues and DynamicArchetype
Log Eigen, OpenCV, and custom C++ types without copying
Log data from Eigen, OpenCV, or your own C++ containers, with or without a copy, using rerun::Collection and CollectionAdapter
Clear out data using tombstones
Mark data invalid after a point in time
Working with MCAP
Open and inspect MCAP files in Rerun
Working with LeRobot datasets
Open and convert LeRobot datasets in Rerun
Optimize chunk count
Reduce chunk count to improve ingest and query performance
Work with .rrd files
Inspect, migrate, trim, split, merge, compact, and append to .rrd files with the rerun rrd command
Loading URDF models
Load and visualize robot models defined in URDF