|
"""Isolet dataset.""" |
|
|
|
from typing import List |
|
|
|
import datasets |
|
|
|
import pandas |
|
|
|
|
|
VERSION = datasets.Version("1.0.0") |
|
|
|
DESCRIPTION = "Isolet dataset from the UCI ML repository." |
|
_HOMEPAGE = "https://archive.ics.uci.edu/ml/datasets/Isolet" |
|
_URLS = ("https://archive-beta.ics.uci.edu/dataset/54/isolet") |
|
_CITATION = """ |
|
@misc{misc_isolet_54, |
|
author = {Cole,Ron & Fanty,Mark}, |
|
title = {{ISOLET}}, |
|
year = {1994}, |
|
howpublished = {UCI Machine Learning Repository}, |
|
note = {{DOI}: \\url{10.24432/C51G69}} |
|
}""" |
|
|
|
|
|
urls_per_split = { |
|
"train": "https://huggingface.co/datasets/mstz/isolet/resolve/main/isolet.zip" |
|
} |
|
features_types_per_config = { |
|
"isolet": { |
|
str(i): datasets.Value("float64") for i in range(617) |
|
} |
|
} |
|
features_types_per_config["isolet"]["617"] = datasets.ClassLabel(num_classes=26) |
|
features_per_config = {k: datasets.Features(features_types_per_config[k]) for k in features_types_per_config} |
|
|
|
|
|
class IsoletConfig(datasets.BuilderConfig): |
|
def __init__(self, **kwargs): |
|
super(IsoletConfig, self).__init__(version=VERSION, **kwargs) |
|
self.features = features_per_config[kwargs["name"]] |
|
|
|
|
|
class Isolet(datasets.GeneratorBasedBuilder): |
|
|
|
DEFAULT_CONFIG = "isolet" |
|
BUILDER_CONFIGS = [ |
|
IsoletConfig(name="isolet", |
|
description="Isolet for letter classification."), |
|
] |
|
|
|
|
|
def _info(self): |
|
info = datasets.DatasetInfo(description=DESCRIPTION, citation=_CITATION, homepage=_HOMEPAGE, |
|
features=features_per_config[self.config.name]) |
|
|
|
return info |
|
|
|
def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]: |
|
downloads = dl_manager.download_and_extract(urls_per_split) |
|
|
|
return [ |
|
datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": downloads["train"]}) |
|
] |
|
|
|
def _generate_examples(self, filepath: str): |
|
data = pandas.read_csv(filepath + "/isolet1+2+3+4.data", header=None).infer_objects() |
|
data = self.preprocess(data, config=self.config.name) |
|
|
|
for row_id, row in data.iterrows(): |
|
data_row = dict(row) |
|
data_row["617"] = int(data_row["617"]) |
|
|
|
yield row_id, data_row |
|
|
|
def preprocess(self, data: pandas.DataFrame, config: str = DEFAULT_CONFIG) -> pandas.DataFrame: |
|
data.columns = [str(i) for i in range(618)] |
|
|
|
data = data.astype({"617": "int8"}) |
|
data.loc[:, "617"] = data["617"].apply(lambda x: int(x) - 1) |
|
data = data.astype({str(i): "float64" for i in range(617)}) |
|
|
|
return data |
|
|
