reuters10k.py

import os

import datasets
from sklearn.preprocessing import MinMaxScaler, LabelEncoder, StandardScaler
import numpy as np


# TODO: Add BibTeX citation
# Find for instance the citation on arxiv or on the dataset repo/website
_CITATION = """\
@InProceedings{huggingface:dataset,
title = {A great new dataset},
author={huggingface, Inc.
},
year={2020}
}
"""

class Reuters10K(datasets.GeneratorBasedBuilder):
    """TODO: Short description of my dataset."""

    VERSION = datasets.Version("0.0.1")

    def _info(self):
        return datasets.DatasetInfo(
            description="Reuters10K dataset",
            version=Reuters10K.VERSION,
        )

    def _split_generators(self, dl_manager):
        train_url = "train.npy"
        test_url = "test.npy"

        downloaded_files = dl_manager.download_and_extract({
            "train": train_url, 
            "test": test_url
        })

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "filepath": downloaded_files["train"]
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "filepath": downloaded_files["test"]
                },
            )
        ]

    def _generate_examples(self, filepath):
        """Yields examples."""
        train_dataset = np.load(filepath, allow_pickle=True)

        X_train = train_dataset.item()['data']
        Y_train = train_dataset.item()['label']

        scaler = MinMaxScaler()
        X_train = scaler.fit_transform(X_train)

        # yield "key", {"text": text, "label": label}
        for i, (x, y) in enumerate(zip(X_train, Y_train)):
            yield i, {"features": x, "label": y}