Skip to content

Examples

Each example is a complete calibration run on a released dataset. It shows:

  • the configuration we used
  • the result it produced
  • how far the result is from a reference calibration, where one exists

Pick the one whose rig looks most like yours.

Looking for a specific sensor? See the Sensor wiki. For what a gyro–accel offset does to tools that cannot model it, see Why calibrate t_ga.

Every example has the same sections, so you can compare any two directly:

  1. The rig: the sensors, their rates and what makes the dataset interesting.
  2. Getting the data: the download, and any preparation the bag needs.
  3. Configuration: the full launch file.
  4. Running it: the command, and what it prints.
  5. Result: the calibration it produced.
  6. Validation: the same quantities from an independent source, and the difference. This part makes an example evidence, not just a demonstration.
  7. Reproducibility: the code version, config and data hash you need to get the same numbers.

The most useful thing you can send is a run on a dataset we don’t have.