LightningPose data conversion

Install NeuroConv with the additional dependencies necessary for reading LightningPose data.

pip install "neuroconv[lightningpose]"

Convert LightningPose pose estimation data to NWB using LightningPoseConverter.

>>> from datetime import datetime
>>> from zoneinfo import ZoneInfo
>>> from pathlib import Path
>>> from neuroconv.converters import LightningPoseConverter

>>> folder_path = BEHAVIOR_DATA_PATH / "lightningpose" / "outputs/2023-11-09/10-14-37/video_preds"
>>> file_path = str(folder_path / "test_vid.csv")
>>> original_video_file_path = str(folder_path / "test_vid.mp4")
>>> # The labeled video file path is optional
>>> labeled_video_file_path = str(folder_path / "labeled_videos/test_vid_labeled.mp4")

>>> converter = LightningPoseConverter(file_path=file_path, original_video_file_path=original_video_file_path, labeled_video_file_path=labeled_video_file_path, verbose=False)
>>> metadata = converter.get_metadata()
>>> # For data provenance we add the time zone information to the conversion
>>> session_start_time = metadata["NWBFile"]["session_start_time"]
>>> tzinfo = ZoneInfo("US/Pacific")
>>> metadata["NWBFile"].update(session_start_time=session_start_time.replace(tzinfo=tzinfo))
>>>
>>> # Choose a path for saving the nwb file and run the conversion
>>> converter.run_conversion(nwbfile_path=path_to_save_nwbfile, metadata=metadata)