bug in loading datasets

[ghost]

Hi,
I need to load a dataset, I use these commands:

from datasets import load_dataset
dataset = load_dataset('csv', data_files={'train': 'sick/train.csv',
                                          'test':  'sick/test.csv',
                                          'validation': 'sick/validation.csv'})
print(dataset['validation'])

the dataset in sick/train.csv are simple csv files representing the data. I am getting this error, do you have an idea how I can solve this? thank you @lhoestq

Using custom data configuration default
Downloading and preparing dataset csv/default-61468fc71a743ec1 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /julia/cache_home_2/datasets/csv/default-61468fc71a743ec1/0.0.0/2960f95a26e85d40ca41a230ac88787f715ee3003edaacb8b1f0891e9f04dda2...
Traceback (most recent call last):
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 485, in incomplete_dir
    yield tmp_dir
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 527, in download_and_prepare
    dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 604, in _download_and_prepare
    self._prepare_split(split_generator, **prepare_split_kwargs)
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 959, in _prepare_split
    for key, table in utils.tqdm(generator, unit=" tables", leave=False, disable=not_verbose):
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/tqdm-4.49.0-py3.7.egg/tqdm/std.py", line 1133, in __iter__
    for obj in iterable:
  File "/julia/cache_home_2/modules/datasets_modules/datasets/csv/2960f95a26e85d40ca41a230ac88787f715ee3003edaacb8b1f0891e9f04dda2/csv.py", line 129, in _generate_tables
    for batch_idx, df in enumerate(csv_file_reader):
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/pandas-1.2.0-py3.7-linux-x86_64.egg/pandas/io/parsers.py", line 1029, in __next__
    return self.get_chunk()
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/pandas-1.2.0-py3.7-linux-x86_64.egg/pandas/io/parsers.py", line 1079, in get_chunk
    return self.read(nrows=size)
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/pandas-1.2.0-py3.7-linux-x86_64.egg/pandas/io/parsers.py", line 1052, in read
    index, columns, col_dict = self._engine.read(nrows)
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/pandas-1.2.0-py3.7-linux-x86_64.egg/pandas/io/parsers.py", line 2056, in read
    data = self._reader.read(nrows)
  File "pandas/_libs/parsers.pyx", line 756, in pandas._libs.parsers.TextReader.read
  File "pandas/_libs/parsers.pyx", line 783, in pandas._libs.parsers.TextReader._read_low_memory
  File "pandas/_libs/parsers.pyx", line 827, in pandas._libs.parsers.TextReader._read_rows
  File "pandas/_libs/parsers.pyx", line 814, in pandas._libs.parsers.TextReader._tokenize_rows
  File "pandas/_libs/parsers.pyx", line 1951, in pandas._libs.parsers.raise_parser_error
pandas.errors.ParserError: Error tokenizing data. C error: Expected 1 fields in line 37, saw 2


During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "write_sick.py", line 19, in <module>
    'validation': 'sick/validation.csv'})
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/load.py", line 612, in load_dataset
    ignore_verifications=ignore_verifications,
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 534, in download_and_prepare
    self._save_info()
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/contextlib.py", line 130, in __exit__
    self.gen.throw(type, value, traceback)
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/site-packages/datasets-1.2.0-py3.7.egg/datasets/builder.py", line 491, in incomplete_dir
    shutil.rmtree(tmp_dir)
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/shutil.py", line 498, in rmtree
    onerror(os.rmdir, path, sys.exc_info())
  File "/julia/libs/anaconda3/envs/success/lib/python3.7/shutil.py", line 496, in rmtree
    os.rmdir(path)
OSError: [Errno 39] Directory not empty: '/julia/cache_home_2/datasets/csv/default-61468fc71a743ec1/0.0.0/2960f95a26e85d40ca41a230ac88787f715ee3003edaacb8b1f0891e9f04dda2.incomplete'

[lhoestq]

Looks like an issue with your csv file. Did you use the right delimiter ?
Apparently at line 37 the CSV reader from pandas reads 2 fields instead of 1.

[lhoestq]

Note that you can pass any argument you would pass to pandas.read_csv as kwargs to load_dataset. For example you can do

from datasets import load_dataset
dataset = load_dataset('csv', data_files=data_files, sep="\t")

for example to use a tab separator.

You can see the full list of arguments here: https://github.com/huggingface/datasets/blob/master/src/datasets/packaged_modules/csv/csv.py

(I've not found the list in the documentation though, we definitely must add them !)

[fghg123]

You can try to convert the file to (CSV UTF-8)