comma2k19.py

import datasets
import glob
import os
import numpy as np

NUM_SHARDS = 10
_URLS = [
      f'https://huggingface.co/datasets/commaai/comma2k19/resolve/main/Chunk_{i}.zip' for i in range(1,NUM_SHARDS+1)
    ]

_DESCRIPTION = """\
comma2k19 is a dataset of over 33 hours of commute in California's 280 highway.
This means 2019 segments, 1 minute long each, on a 20km section of highway driving between California's San Jose and San Francisco.
comma2k19 is a fully reproducible and scalable dataset.
The data was collected using comma EONs that has sensors similar to those of any modern smartphone including a road-facing camera, phone GPS, thermometers and 9-axis IMU.
Additionally, the EON captures raw GNSS measurements and all CAN data sent by the car with a comma grey panda.
"""

class Comma2k19(datasets.GeneratorBasedBuilder):

  def _info(self):
    return datasets.DatasetInfo(
      description=_DESCRIPTION,
      features=datasets.Features(
        {"path": datasets.Value("string")}
      )
    )

  def _split_generators(self, dl_manager):
    """Returns SplitGenerators."""
    dl_manager.download_config.ignore_url_params = True
    downloaded_files = dl_manager.download(_URLS)
    local_extracted_archive = dl_manager.extract(downloaded_files) if not dl_manager.is_streaming else [None]*len(downloaded_files)
    return [
      datasets.SplitGenerator(
        name=str(i),
        gen_kwargs={"local_extracted_archive":local_extracted_archive[i], "files": dl_manager.iter_archive(downloaded_files[i])}
        ) for i in range(len(downloaded_files))]

  def _generate_examples(self, local_extracted_archive, files):
    files = [os.path.join(dp, f) for dp, dn, filenames in os.walk(local_extracted_archive) for f in filenames]
    for path in files:
      yield path, {'path': path}

  def _get_examples_iterable_for_split(self, split_generator):
    for path in split_generator.gen_kwargs['files']:
      yield path[0], {'path': path[0]}