This directory contains tools for recording, storing, and analyzing vehicle path data.
Contains the raw and processed data from vehicle runs.
original_run/: Data from the initial/reference run.smoothed_run/: Data from runs using the smoothed path.output.pickle: The graph data generated by the mapping tool.
Scripts related to data acquisition. See the Recording Documentation for detailed instructions.
record_ros.py: A ROS node script to record vehicle pose (/ndt_pose) and steering (/mpc/steer_angle) to CSV files.
Compares two graph pickle files (output.pickle) to analyze differences in node positions and attributes (Yaw, etc.).
python compare_pickles.py path/to/run1/output.pickle path/to/run2_folderpython analysis/compare_pickles.py analysis/recorded_data/original_run/output.pickle analysis/recorded_data/smoothed_run/output.pickleCompares recorded CSV vehicle data (trajectory, steering, speed) between two runs.
python compare_runs.py recorded_data/run1 recorded_data/run2python analysis/compare_runs.py analysis/recorded_data/original_run analysis/recorded_data/smoothed_runAnalyzes lane data from .npy files.
Note: This script currently uses hardcoded paths in
__main__. You may need to edit it to point to your specific data files.
Additional analysis utilities for debugging vehicle behavior.
Note: Requires manual configuration of input file paths within the script.