pums.py

"""Pums Dataset"""

from typing import List
from functools import partial

import datasets

import pandas


VERSION = datasets.Version("1.0.0")

_ENCODING_DICS = {
	"class": {
		"- 50000.": 0,
		"50000+.": 1
	}
}
_BASE_FEATURE_NAMES = [
	"age",
	"class_of_worker",
	"detailed_industry_recode",
	"detailed_occupation_recode",
	"education",
	"wage_per_hour",
	"enroll_in_edu_inst_last_wk",
	"marital_stat",
	"major_industry_code",
	"major_occupation_code",
	"race",
	"hispanic_origin",
	"sex",
	"member_of_a_labor_union",
	"reason_for_unemployment",
	"full_or_part_time_employment_stat",
	"capital_gains",
	"capital_losses",
	"dividends_from_stocks",
	"tax_filer_stat",
	"region_of_previous_residence",
	"state_of_previous_residence",
	"detailed_household_and_family_stat",
	"detailed_household_summary_in_household",
	"migration_code_change_in_msa",
	"migration_code_change_in_reg",
	"migration_code_move_within_reg",
	"live_in_this_house_1_year_ago",
	"migration_prev_res_in_sunbelt",
	"num_persons_worked_for_employer",
	"family_members_under_18",
	"country_of_birth_father",
	"country_of_birth_mother",
	"country_of_birth_self",
	"citizenship",
	"own_business_or_self_employed",
	"fill_inc_questionnaire_for_veteran_admin",
	"veterans_benefits",
	"weeks_worked_in_year",
	"year",
	"class",
]

DESCRIPTION = "Pums dataset."
_HOMEPAGE = "https://archive-beta.ics.uci.edu/dataset/116/us+census+data+1990"
_URLS = ("https://archive-beta.ics.uci.edu/dataset/116/us+census+data+1990")
_CITATION = """
@misc{misc_us_census_data_(1990)_116,
  author       = {Meek,Meek, Thiesson,Thiesson & Heckerman,Heckerman},
  title        = {{US Census Data (1990)}},
  howpublished = {UCI Machine Learning Repository},
  note         = {{DOI}: \\url{10.24432/C5VP42}}
}
"""

# Dataset info
urls_per_split = {
	"train": "https://huggingface.co/datasets/mstz/pums/resolve/main/pums.csv"
}
features_types_per_config = {
	"pums": {
		"age": datasets.Value("int64"),
		"class_of_worker": datasets.Value("string"),
		"detailed_industry_recode": datasets.Value("string"),
		"detailed_occupation_recode": datasets.Value("string"),
		"education": datasets.Value("string"),
		"wage_per_hour": datasets.Value("int64"),
		"enroll_in_edu_inst_last_wk": datasets.Value("string"),
		"marital_stat": datasets.Value("string"),
		"major_industry_code": datasets.Value("string"),
		"major_occupation_code": datasets.Value("string"),
		"race": datasets.Value("string"),
		"hispanic_origin": datasets.Value("string"),
		"sex": datasets.Value("string"),
		"member_of_a_labor_union": datasets.Value("string"),
		"reason_for_unemployment": datasets.Value("string"),
		"full_or_part_time_employment_stat": datasets.Value("string"),
		"capital_gains": datasets.Value("int64"),
		"capital_losses": datasets.Value("int64"),
		"dividends_from_stocks": datasets.Value("int64"),
		"tax_filer_stat": datasets.Value("string"),
		"region_of_previous_residence": datasets.Value("string"),
		"state_of_previous_residence": datasets.Value("string"),
		"detailed_household_and_family_stat": datasets.Value("string"),
		"detailed_household_summary_in_household": datasets.Value("string"),
		"migration_code_change_in_msa": datasets.Value("string"),
		"migration_code_change_in_reg": datasets.Value("string"),
		"migration_code_move_within_reg": datasets.Value("string"),
		"live_in_this_house_1_year_ago": datasets.Value("string"),
		"migration_prev_res_in_sunbelt": datasets.Value("string"),
		"num_persons_worked_for_employer": datasets.Value("int64"),
		"family_members_under_18": datasets.Value("string"),
		"country_of_birth_father": datasets.Value("string"),
		"country_of_birth_mother": datasets.Value("string"),
		"country_of_birth_self": datasets.Value("string"),
		"citizenship": datasets.Value("string"),
		"own_business_or_self_employed": datasets.Value("int64"),
		"fill_inc_questionnaire_for_veteran_admin": datasets.Value("string"),
		"veterans_benefits": datasets.Value("int64"),
		"weeks_worked_in_year": datasets.Value("int64"),
		"year": datasets.Value("int64"),
		"class": datasets.ClassLabel(num_classes=2)
	}
}
features_per_config = {k: datasets.Features(features_types_per_config[k]) for k in features_types_per_config}


class PumsConfig(datasets.BuilderConfig):
	def __init__(self, **kwargs):
		super(PumsConfig, self).__init__(version=VERSION, **kwargs)
		self.features = features_per_config[kwargs["name"]]


class Pums(datasets.GeneratorBasedBuilder):
	# dataset versions
	DEFAULT_CONFIG = "pums"
	BUILDER_CONFIGS = [PumsConfig(name="pums", description="Pums for binary 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, header=None)
		data = self.preprocess(data)

		for row_id, row in data.iterrows():
			data_row = dict(row)

			yield row_id, data_row

	def preprocess(self, data: pandas.DataFrame) -> pandas.DataFrame:
		data.columns = _BASE_FEATURE_NAMES

		for feature in _ENCODING_DICS:
			encoding_function = partial(self.encode, feature)
			data.loc[:, feature] = data[feature].apply(encoding_function)
		
		data = data.rename(columns={"instance migration_code_change_in_msa": "migration_code_change_in_msa"})
				
		return data[list(features_types_per_config[self.config.name].keys())]

	def encode(self, feature, value):
		if feature in _ENCODING_DICS:
			return _ENCODING_DICS[feature][value]
		raise ValueError(f"Unknown feature: {feature}")