OptimizeCalendar.py

import numpy as np
import pandas as pd
import time
import os
from dotenv import load_dotenv

load_dotenv('./dev.env')

DatasetsFolderPath = os.getenv("DATASETS_FOLDER_PATTH")

'''
Create the result for period grouping
'''
def groupDate(group):
    return pd.Series([group.iloc[0]['date'], group.iloc[-1]['date'], group.shape[0],  group.iloc[0]['minimum_nights'],  group.iloc[0]['maximum_nights']])
'''
Import the calendar dataset and group by open/closed period to remove useless data
'''
def OptimizeCalendar(filename):
    calendar = pd.read_csv(f'{DatasetsFolderPath}/calendar/calendar-{filename}.csv',sep=",")

    calendar = calendar.sort_values(by=["listing_id", "date"])
    calendar = calendar.reset_index()
    # drop useless tables
    calendar = calendar.drop(list(set(['price','index','adjusted_price']) & set(calendar.columns)),axis=1)

    if 'minimum_nights' not in calendar.columns:
        calendar['minimum_nights'] = 0
    if 'maximum_nights' not in calendar.columns:
        calendar['maximum_nights'] = 1125

    # Group periods
    adj_check = (calendar.available != calendar.available.shift()).cumsum() # Avoid to grop no consecutive rows
    newData = calendar.groupby(['listing_id','available',adj_check], as_index=False, sort=False).apply(groupDate)
    newData = newData.rename({0:'start_date',1:'end_date',2:'num_day',3:'minimum_nights',4:'maximum_nights'},axis=1)

    return newData

def ProcessAndSave(fileNameDate,SavedName):
    exists = os.path.isfile(f"{DatasetsFolderPath}/saved/{fileNameDate}/{SavedName}-{fileNameDate}.csv") 
    if exists:
        print(f'--- Used {DatasetsFolderPath}/saved/{fileNameDate}/{SavedName}-{fileNameDate}.csv ---')
        return pd.read_csv(f"{DatasetsFolderPath}/saved/{fileNameDate}/{SavedName}-{fileNameDate}.csv",sep=",")
    else:
        start_time = time.time()
        df = OptimizeCalendar(fileNameDate)
        print(f'--- Optimizing {fileNameDate} : {time.time() - start_time} ---')
        df.to_csv(f"{DatasetsFolderPath}/saved/{fileNameDate}/{SavedName}-{fileNameDate}.csv", index = False)
        return df



if __name__ == "__main__":
    
    start_time = time.time()
    newData = OptimizeCalendar("calendar-2017-01")
    print("---  %s seconds ---" % (time.time() - start_time))