generate_fake_data.py

import random
from typing import Generator

import pandas as pd
from faker import Faker


def generate_fake_data(
    func_name: str,
    data_length: int = 10000,
    data_types: int = 100,
    locale: str = "ja_JP",
) -> Generator[str, None, None]:
    """
    :func_name: 使用するFakerインスタンスのproviderの名前
    :data_length: データ件数
    :data_types: データの種類の上限数
    :locale: Fakerインスタンスのlocale
    """
    faker = Faker(locale)
    faker_func = getattr(faker, func_name)

    faker_types = [faker_func() for _ in range(data_types)]
    return (faker_types[random.randrange(0, data_types)] for _ in range(data_length))