숙소 더미데이터 생성

create_dummy_data_rooms_images.py

import requests
import random
import numpy as np
import pandas as pd
from scipy.stats import skewnorm 
from faker import Faker



def create_rooms(room_size, host_id_size):
    fake = Faker('ko-KR')
    df_rooms = pd.DataFrame(columns=['host_id', 'title', 'description', 'address' , 'lat', 'lng' ,
                                'bathroom_count', 'bed_count', 'bedroom_count', 'header_count_capacity',
                                'cleaning_fee', 'daily_price', 'lodging_tax_ratio', 'sale_ratio', 'service_fee',
                                'rating_star_score', 'review_count'])
    
    room_types = ['호텔', '민박', '모텔', '도미토리', '아파트', '멘션', '게스트하우스', '한옥']

    skew_norm_distribution = skewnorm.rvs(20, size=room_size + 300)
    price_distribution = np.round((skew_norm_distribution[skew_norm_distribution > 0] * 300000) + 3000, 1)
    for row_i in range(room_size):
        lat_lng = fake.local_latlng(country_code= 'KR') # ('37.1759', '128.9889', 'T’aebaek', 'KR', 'Asia/Seoul')

        host_id = random.randint(1, host_id_size)
        title = fake.company() + ' ' + np.random.choice(room_types) # 숙소 이름 e.g. 김김이 게스트하우스
        description = fake.text() # 숙소 설명 e.g. Quibusdam voluptatem omnis odio. Veniam nihil amet. Ratione minus repudiandae enim accusamus possimus.
        address = fake.address() # 
        lat = lat_lng[0] # 위도 e.g 34.8825
        lng = lat_lng[1] # 경도 e.g 128.62667
        bathroom_count = random.randint(1, 3) # 화장실 수 e.g. 3
        bed_count = random.randint(1, 5) # 침대 수 e.g. 2
        bedroom_count = random.randint(1, 5) # 침실 수 e.g. 2
        headcount_capacity = random.randint(1, 10) # 최대 인원수 e.g. 8
        cleaning_fee = random.randint(10, 100) * 100 # 청소비 e.g. 5000
        daily_price = int(price_distribution[row_i]) # 일일 가격 e.g. 30000 ~ 1500000 편향분포
        lodging_tax_ratio = 10 # 세금(10%) e.g. 10
        sale_ratio = random.randint(1, 10) # 할인률(10%) e.g. 10
        service_fee = random.randint(5, 50) * 100 # 서비스 비용 e.g. 5000
        rating_star_score = round(random.uniform(1, 5), 2) # 평점 e.g. 3.3 
        review_count = random.randint(10, 500) # 리뷰개수

        df_rooms.loc[row_i] = [host_id, title, description, address, lat, lng,
                                bathroom_count, bed_count, bedroom_count, headcount_capacity,
                                cleaning_fee, daily_price, lodging_tax_ratio, sale_ratio, service_fee,
                                rating_star_score, review_count]
    return df_rooms


def create_room_images(room_size, random_image_size = 100, image_size_row = 400, image_size_col = 400):
    
    df_room_image = pd.DataFrame(columns=['room_id', 'image_url'])

    room_image_list = []
    for _ in range(1, random_image_size):
        image_url = 'https://picsum.photos/' + str(image_size_row) + '/' + str(image_size_col) + '?random=1'
        r = requests.get(image_url, allow_redirects=False)
        room_image_list.append(str(r.headers['Location']))

    global_image_i = 0
    for room_i in range(1, room_size + 1):        
        for _ in range(random.randint(1, 7)):
            df_room_image.loc[global_image_i] = [room_i, random.choice(room_image_list)]
            global_image_i = global_image_i + 1
            
    return df_room_image
            
df_rooms = create_rooms(5000, 2000) # 5000개의 숙소 데이터 생성, FK host id는 2000 까지 있다고 가정
df_room_image = create_room_images(500) 
#df_room_image = create_room_images(500, 100, 400, 400) # 랜덤 이미지개수, 사이즈 조정 가능 (400x400)

df_rooms.to_csv('./dummy_rooms.csv', index = False)
df_room_image.to_csv('./dummy_room_images.csv', index = False)