"""Speeddating Dataset""" from typing import List from functools import partial import datasets import pandas VERSION = datasets.Version("1.0.0") _BASE_FEATURE_NAMES = [ "dater_gender", "dater_age", "dated_age", "age_difference", "dater_race", "dated_race", "are_same_race", "same_race_importance_for_dater", "same_religion_importance_for_dater", "attractiveness_importance_for_dated", "sincerity_importance_for_dated", "intelligence_importance_for_dated", "humor_importance_for_dated", "ambition_importance_for_dated", "shared_interests_importance_for_dated", "attractiveness_score_of_dater_from_dated", "sincerity_score_of_dater_from_dated", "intelligence_score_of_dater_from_dated", "humor_score_of_dater_from_dated", "ambition_score_of_dater_from_dated", "shared_interests_score_of_dater_from_dated", "attractiveness_importance_for_dater", "sincerity_importance_for_dater", "intelligence_importance_for_dater", "humor_importance_for_dater", "ambition_importance_for_dater", "shared_interests_importance_for_dater", "self_reported_attractiveness_of_dater", "self_reported_sincerity_of_dater", "self_reported_intelligence_of_dater", "self_reported_humor_of_dater", "self_reported_ambition_of_dater", "reported_attractiveness_of_dated_from_dater", "reported_sincerity_of_dated_from_dater", "reported_intelligence_of_dated_from_dater", "reported_humor_of_dated_from_dater", "reported_ambition_of_dated_from_dater", "reported_shared_interests_of_dated_from_dater", "dater_interest_in_sports", "dater_interest_in_tvsports", "dater_interest_in_exercise", "dater_interest_in_dining", "dater_interest_in_museums", "dater_interest_in_art", "dater_interest_in_hiking", "dater_interest_in_gaming", "dater_interest_in_clubbing", "dater_interest_in_reading", "dater_interest_in_tv", "dater_interest_in_theater", "dater_interest_in_movies", "dater_interest_in_concerts", "dater_interest_in_music", "dater_interest_in_shopping", "dater_interest_in_yoga", "interests_correlation", "expected_satisfaction_of_dater", "expected_number_of_likes_of_dater_from_20_people", "expected_number_of_dates_for_dater", "dater_liked_dated", "probability_dated_wants_to_date", "already_met_before", "dater_wants_to_date", "dated_wants_to_date", "is_match" ] _ENCODING_DICS = { "sex": { "female": 0, "male": 1 } } DESCRIPTION = "Speed-dating dataset." _HOMEPAGE = "https://www.openml.org/search?type=data&sort=nr_of_likes&status=active&id=40536" _URLS = ("https://huggingface.co/datasets/mstz/speeddating/raw/main/speeddating.csv") _CITATION = """""" # Dataset info urls_per_split = { "train": "https://huggingface.co/datasets/mstz/speeddating/raw/main/speeddating.csv", } features_types_per_config = { "dating": { "dater_gender": datasets.Value("int8"), "dater_age": datasets.Value("int8"), "dated_age": datasets.Value("int8"), "age_difference": datasets.Value("int8"), "dater_race": datasets.Value("string"), "dated_race": datasets.Value("string"), "are_same_race": datasets.Value("int8"), "same_race_importance_for_dater": datasets.Value("int8"), "same_religion_importance_for_dater": datasets.Value("int8"), "attractiveness_importance_for_dated": datasets.Value("int8"), "sincerity_importance_for_dated": datasets.Value("int8"), "intelligence_importance_for_dated": datasets.Value("int8"), "humor_importance_for_dated": datasets.Value("int8"), "ambition_importance_for_dated": datasets.Value("int8"), "shared_interests_importance_for_dated": datasets.Value("int8"), "attractiveness_score_of_dater_from_dated": datasets.Value("int8"), "sincerity_score_of_dater_from_dated": datasets.Value("int8"), "intelligence_score_of_dater_from_dated": datasets.Value("int8"), "humor_score_of_dater_from_dated": datasets.Value("int8"), "ambition_score_of_dater_from_dated": datasets.Value("int8"), "shared_interests_score_of_dater_from_dated": datasets.Value("int8"), "attractiveness_importance_for_dater": datasets.Value("int8"), "sincerity_importance_for_dater": datasets.Value("int8"), "intelligence_importance_for_dater": datasets.Value("int8"), "humor_importance_for_dater": datasets.Value("int8"), "ambition_importance_for_dater": datasets.Value("int8"), "shared_interests_importance_for_dater": datasets.Value("int8"), "self_reported_attractiveness_of_dater": datasets.Value("int8"), "self_reported_sincerity_of_dater": datasets.Value("int8"), "self_reported_intelligence_of_dater": datasets.Value("int8"), "self_reported_humor_of_dater": datasets.Value("int8"), "self_reported_ambition_of_dater": datasets.Value("int8"), "reported_attractiveness_of_dated_from_dater": datasets.Value("int8"), "reported_sincerity_of_dated_from_dater": datasets.Value("int8"), "reported_intelligence_of_dated_from_dater": datasets.Value("int8"), "reported_humor_of_dated_from_dater": datasets.Value("int8"), "reported_ambition_of_dated_from_dater": datasets.Value("int8"), "reported_shared_interests_of_dated_from_dater": datasets.Value("int8"), "dater_interest_in_sports": datasets.Value("int8"), "dater_interest_in_tvsports": datasets.Value("int8"), "dater_interest_in_exercise": datasets.Value("int8"), "dater_interest_in_dining": datasets.Value("int8"), "dater_interest_in_museums": datasets.Value("int8"), "dater_interest_in_art": datasets.Value("int8"), "dater_interest_in_hiking": datasets.Value("int8"), "dater_interest_in_gaming": datasets.Value("int8"), "dater_interest_in_clubbing": datasets.Value("int8"), "dater_interest_in_reading": datasets.Value("int8"), "dater_interest_in_tv": datasets.Value("int8"), "dater_interest_in_theater": datasets.Value("int8"), "dater_interest_in_movies": datasets.Value("int8"), "dater_interest_in_concerts": datasets.Value("int8"), "dater_interest_in_music": datasets.Value("int8"), "dater_interest_in_shopping": datasets.Value("int8"), "dater_interest_in_yoga": datasets.Value("int8"), "interests_correlation": datasets.Value("float16"), "expected_satisfaction_of_dater": datasets.Value("int8"), "expected_number_of_likes_of_dater_from_20_people": datasets.Value("int8"), "expected_number_of_dates_for_dater": datasets.Value("int8"), "dater_liked_dated": datasets.Value("int8"), "probability_dated_wants_to_date": datasets.Value("int8"), "already_met_before": datasets.Value("int8"), "dater_wants_to_date": datasets.Value("int8"), "dated_wants_to_date": datasets.Value("int8"), "is_match": datasets.ClassLabel(num_classes=2, names=("no", "yes")) } } features_per_config = {k: datasets.Features(features_types_per_config[k]) for k in features_types_per_config} class SpeeddatingConfig(datasets.BuilderConfig): def __init__(self, **kwargs): super(SpeeddatingConfig, self).__init__(version=VERSION, **kwargs) self.features = features_per_config[kwargs["name"]] class Speeddating(datasets.GeneratorBasedBuilder): # dataset versions DEFAULT_CONFIG = "dating" BUILDER_CONFIGS = [ SpeeddatingConfig(name="dating", description="Binary classification."), ] def _info(self): if self.config.name not in features_per_config: raise ValueError(f"Unknown configuration: {self.config.name}") 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) data = self.preprocess(data, config=self.config.name) for row_id, row in data.iterrows(): data_row = dict(row) yield row_id, data_row def preprocess(self, data: pandas.DataFrame, config: str = "dating") -> pandas.DataFrame: data.loc[data.race == "?", "race"] = "unknown" data.loc[data.race == "Asian/Pacific Islander/Asian-American", "race"] = "asian" data.loc[data.race == "European/Caucasian-American", "race"] = "caucasian" data.loc[data.race == "Other", "race"] = "other" data.loc[data.race == "Latino/Hispanic American", "race"] = "hispanic" data.loc[data.race == "Black/African American", "race"] = "african-american" sex_transform = partial(self.encoding_dics, "sex") data.loc[:, "gender"] = data.gender.apply(sex_transform) data = data.rename(columns={"gender": "sex"}) data.drop("has_null", axis="columns", inplace=True) data.drop("field", axis="columns", inplace=True) data.drop("wave", axis="columns", inplace=True) # data.drop("d_age", axis="columns", inplace=True) data.drop("d_d_age", axis="columns", inplace=True) data.drop("d_importance_same_race", axis="columns", inplace=True) data.drop("d_importance_same_religion", axis="columns", inplace=True) data.drop("d_pref_o_attractive", axis="columns", inplace=True) data.drop("d_pref_o_sincere", axis="columns", inplace=True) data.drop("d_pref_o_intelligence", axis="columns", inplace=True) data.drop("d_pref_o_funny", axis="columns", inplace=True) data.drop("d_pref_o_ambitious", axis="columns", inplace=True) data.drop("d_pref_o_shared_interests", axis="columns", inplace=True) data.drop("d_attractive_o", axis="columns", inplace=True) data.drop("d_sinsere_o", axis="columns", inplace=True) data.drop("d_intelligence_o", axis="columns", inplace=True) data.drop("d_funny_o", axis="columns", inplace=True) data.drop("d_ambitous_o", axis="columns", inplace=True) data.drop("d_shared_interests_o", axis="columns", inplace=True) data.drop("d_attractive_important", axis="columns", inplace=True) data.drop("d_sincere_important", axis="columns", inplace=True) data.drop("d_intellicence_important", axis="columns", inplace=True) data.drop("d_funny_important", axis="columns", inplace=True) data.drop("d_ambtition_important", axis="columns", inplace=True) data.drop("d_shared_interests_important", axis="columns", inplace=True) data.drop("d_attractive", axis="columns", inplace=True) data.drop("d_sincere", axis="columns", inplace=True) data.drop("d_intelligence", axis="columns", inplace=True) data.drop("d_funny", axis="columns", inplace=True) data.drop("d_ambition", axis="columns", inplace=True) data.drop("d_attractive_partner", axis="columns", inplace=True) data.drop("d_sincere_partner", axis="columns", inplace=True) data.drop("d_intelligence_partner", axis="columns", inplace=True) data.drop("d_funny_partner", axis="columns", inplace=True) data.drop("d_ambition_partner", axis="columns", inplace=True) data.drop("d_shared_interests_partner", axis="columns", inplace=True) data.drop("d_sports", axis="columns", inplace=True) data.drop("d_tvsports", axis="columns", inplace=True) data.drop("d_exercise", axis="columns", inplace=True) data.drop("d_dining", axis="columns", inplace=True) data.drop("d_museums", axis="columns", inplace=True) data.drop("d_art", axis="columns", inplace=True) data.drop("d_hiking", axis="columns", inplace=True) data.drop("d_gaming", axis="columns", inplace=True) data.drop("d_clubbing", axis="columns", inplace=True) data.drop("d_reading", axis="columns", inplace=True) data.drop("d_tv", axis="columns", inplace=True) data.drop("d_theater", axis="columns", inplace=True) data.drop("d_movies", axis="columns", inplace=True) data.drop("d_concerts", axis="columns", inplace=True) data.drop("d_music", axis="columns", inplace=True) data.drop("d_shopping", axis="columns", inplace=True) data.drop("d_yoga", axis="columns", inplace=True) data.drop("d_interests_correlate", axis="columns", inplace=True) data.drop("d_expected_happy_with_sd_people", axis="columns", inplace=True) data.drop("d_expected_num_interested_in_me", axis="columns", inplace=True) data.drop("d_expected_num_matches", axis="columns", inplace=True) data.drop("d_like", axis="columns", inplace=True) data.drop("d_guess_prob_liked", axis="columns", inplace=True) data = data[data.age != "?"] data = data[data.importance_same_race != "?"] data = data[data.pref_o_attractive != "?"] data = data[data.pref_o_sincere != "?"] data = data[data.interests_correlate != "?"] print(data.columns) print(data.head()) data.columns = _BASE_FEATURE_NAMES if config == "dating": return data else: raise ValueError(f"Unknown config: {config}") def encoding_dics(self, feature, value): if feature in _ENCODING_DICS: return _ENCODING_DICS[feature][value] raise ValueError(f"Unknown feature: {feature}")