Convert dataset to Parquet

README.md

@@ -23,8 +23,16 @@ dataset_info:
   - name: test
     num_bytes: 72971
     num_examples: 498
-  download_size: 81363704
+  download_size: 122740838
   dataset_size: 224615661
+configs:
+- config_name: sentiment140
+  data_files:
+  - split: train
+    path: sentiment140/train-*
+  - split: test
+    path: sentiment140/test-*
+  default: true
 train-eval-index:
 - config: sentiment140
   task: text-classification

sentiment140.py

@@ -1,115 +0,0 @@
-import csv
-import os
-
-import datasets
-
-
-_CITATION = """\
-@article{go2009twitter,
-  title={Twitter sentiment classification using distant supervision},
-  author={Go, Alec and Bhayani, Richa and Huang, Lei},
-  journal={CS224N project report, Stanford},
-  volume={1},
-  number={12},
-  pages={2009},
-  year={2009}
-}
-"""
-
-_DESCRIPTION = """\
-Sentiment140 consists of Twitter messages with emoticons, which are used as noisy labels for
-sentiment classification. For more detailed information please refer to the paper.
-"""
-_URL = "http://help.sentiment140.com/home"
-_DATA_URL = "https://cs.stanford.edu/people/alecmgo/trainingandtestdata.zip"
-
-_TEST_FILE_NAME = "testdata.manual.2009.06.14.csv"
-_TRAIN_FILE_NAME = "training.1600000.processed.noemoticon.csv"
-
-
-class Sentiment140Config(datasets.BuilderConfig):
-
-    """BuilderConfig for Break"""
-
-    def __init__(self, data_url, **kwargs):
-        """BuilderConfig for BlogAuthorship
-
-        Args:
-          data_url: `string`, url to the dataset (word or raw level)
-          **kwargs: keyword arguments forwarded to super.
-        """
-        super(Sentiment140Config, self).__init__(version=datasets.Version("1.0.0", ""), **kwargs)
-        self.data_url = data_url
-
-
-class Sentiment140(datasets.GeneratorBasedBuilder):
-
-    VERSION = datasets.Version("0.1.0")
-    BUILDER_CONFIGS = [
-        Sentiment140Config(
-            name="sentiment140",
-            data_url=_DATA_URL,
-            description="sentiment classification dataset. Twitter messages are classified as either 'positive'=0, 'neutral'=1 or 'negative'=2.",
-        )
-    ]
-
-    def _info(self):
-        return datasets.DatasetInfo(
-            # This is the description that will appear on the datasets page.
-            description=_DESCRIPTION,
-            # datasets.features.FeatureConnectors
-            features=datasets.Features(
-                {
-                    "text": datasets.Value("string"),
-                    "date": datasets.Value("string"),
-                    "user": datasets.Value("string"),
-                    "sentiment": datasets.Value("int32"),
-                    "query": datasets.Value("string"),
-                }
-            ),
-            # If there's a common (input, target) tuple from the features,
-            # specify them here. They'll be used if as_supervised=True in
-            # builder.as_dataset.
-            supervised_keys=None,
-            # Homepage of the dataset for documentation
-            homepage=_URL,
-            citation=_CITATION,
-        )
-
-    def _split_generators(self, dl_manager):
-        """Returns SplitGenerators."""
-        data_dir = dl_manager.download_and_extract(_DATA_URL)
-
-        test_csv_file = os.path.join(data_dir, _TEST_FILE_NAME)
-        train_csv_file = os.path.join(data_dir, _TRAIN_FILE_NAME)
-
-        if self.config.name == "sentiment140":
-            return [
-                datasets.SplitGenerator(
-                    name=datasets.Split.TRAIN,
-                    # These kwargs will be passed to _generate_examples
-                    gen_kwargs={"file_path": train_csv_file},
-                ),
-                datasets.SplitGenerator(
-                    name=datasets.Split.TEST,
-                    # These kwargs will be passed to _generate_examples
-                    gen_kwargs={"file_path": test_csv_file},
-                ),
-            ]
-        else:
-            raise NotImplementedError(f"{self.config.name} does not exist")
-
-    def _generate_examples(self, file_path):
-        """Yields examples."""
-
-        with open(file_path, encoding="ISO-8859-1") as f:
-            data = csv.reader(f, delimiter=",", quotechar='"')
-            for row_id, row in enumerate(data):
-                sentiment, tweet_id, date, query, user_name, message = row
-                yield f"{row_id}_{tweet_id}", {
-                    "text": message,
-                    "date": date,
-                    "user": user_name,
-                    "sentiment": int(sentiment),
-                    "query": query,
-                }

sentiment140/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ab0d3f4d21d98b990b393d85988d7b651857600657c26693dfe38790d8114e0
+size 45576

sentiment140/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc80f88c1eb49b903fc7d725f5f08246635da7a4102962d8265bff2a29f1ad5e
+size 122695262