TravelPlannerLeaderboard / utils /budget_estimation.py
hsaest's picture
Upload folder using huggingface_hub
6159f52 verified
raw history blame
No virus
8.82 kB
from tools.accommodations.apis import Accommodations
from tools.flights.apis import Flights
from tools.restaurants.apis import Restaurants
from tools.rank.apis import Rank
from tools.filter.apis import Filter
from tools.googleDistanceMatrix.apis import GoogleDistanceMatrix
import pandas as pd
hotel = Accommodations()
flight = Flights()
flight.load_db()
restaurant = Restaurants()
rank = Rank()
filter = Filter()
distanceMatrix = GoogleDistanceMatrix()
def estimate_budget(data, mode):
"""
Estimate the budget based on the mode (lowest, highest, average) for flight, hotel, or restaurant data.
"""
if mode == "lowest":
return min(data)
elif mode == "highest":
return max(data)
elif mode == "average":
# filter the nan values
data = [x for x in data if str(x) != 'nan']
return sum(data) / len(data)
def budget_calc(org, dest, days, date:list , people_number=None, local_constraint = None):
"""
Calculate the estimated budget for all three modes: lowest, highest, average.
grain: city, state
"""
if days == 3:
grain = "city"
elif days in [5,7]:
grain = "state"
if grain not in ["city", "state"]:
raise ValueError("grain must be one of city, state")
# Multipliers based on days
multipliers = {
3: {"flight": 2, "hotel": 3, "restaurant": 9},
5: {"flight": 3, "hotel": 5, "restaurant": 15},
7: {"flight": 4, "hotel": 7, "restaurant": 21}
}
if grain == "city":
hotel_data = hotel.run(dest)
restaurant_data = restaurant.run(dest)
flight_data = flight.data[(flight.data["DestCityName"] == dest) & (flight.data["OriginCityName"] == org)]
elif grain == "state":
city_set = open('/home/user/app/database/background/citySet_with_states.txt').read().strip().split('\n')
all_hotel_data = []
all_restaurant_data = []
all_flight_data = []
for city in city_set:
if dest == city.split('\t')[1]:
candidate_city = city.split('\t')[0]
# Fetch data for the current city
current_hotel_data = hotel.run(candidate_city)
current_restaurant_data = restaurant.run(candidate_city)
current_flight_data = flight.data[(flight.data["DestCityName"] == candidate_city) & (flight.data["OriginCityName"] == org)]
# Append the dataframes to the lists
all_hotel_data.append(current_hotel_data)
all_restaurant_data.append(current_restaurant_data)
all_flight_data.append(current_flight_data)
# Use concat to combine all dataframes in the lists
hotel_data = pd.concat(all_hotel_data, axis=0)
restaurant_data = pd.concat(all_restaurant_data, axis=0)
flight_data = pd.concat(all_flight_data, axis=0)
# flight_data should be in the range of supported date
flight_data = flight_data[flight_data['FlightDate'].isin(date)]
if people_number:
hotel_data = hotel_data[hotel_data['maximum occupancy'] >= people_number]
if local_constraint:
if local_constraint['transportation'] == 'no self-driving':
if grain == "city":
if len(flight_data[flight_data['FlightDate'] == date[0]]) < 2:
raise ValueError("No flight data available for the given constraints.")
elif grain == "state":
if len(flight_data[flight_data['FlightDate'] == date[0]]) < 10:
raise ValueError("No flight data available for the given constraints.")
elif local_constraint['transportation'] == 'no flight':
if len(flight_data[flight_data['FlightDate'] == date[0]]) < 2 or flight_data.iloc[0]['Distance'] > 800:
raise ValueError("Impossible")
# if local_constraint['flgiht time']:
# if local_constraint['flgiht time'] == 'morning':
# flight_data = flight_data[flight_data['DepTime'] < '12:00']
# elif local_constraint['flgiht time'] == 'afternoon':
# flight_data = flight_data[(flight_data['DepTime'] >= '12:00') & (flight_data['DepTime'] < '18:00')]
# elif local_constraint['flgiht time'] == 'evening':
# flight_data = flight_data[flight_data['DepTime'] >= '18:00']
if local_constraint['room type']:
if local_constraint['room type'] == 'shared room':
hotel_data = hotel_data[hotel_data['room type'] == 'Shared room']
elif local_constraint['room type'] == 'not shared room':
hotel_data = hotel_data[(hotel_data['room type'] == 'Private room') | (hotel_data['room type'] == 'Entire home/apt')]
elif local_constraint['room type'] == 'private room':
hotel_data = hotel_data[hotel_data['room type'] == 'Private room']
elif local_constraint['room type'] == 'entire room':
hotel_data = hotel_data[hotel_data['room type'] == 'Entire home/apt']
if days == 3:
if len(hotel_data) < 3:
raise ValueError("No hotel data available for the given constraints.")
elif days == 5:
if len(hotel_data) < 5:
raise ValueError("No hotel data available for the given constraints.")
elif days == 7:
if len(hotel_data) < 7:
raise ValueError("No hotel data available for the given constraints.")
if local_constraint['house rule']:
if local_constraint['house rule'] == 'parties':
# the house rule should not contain 'parties'
hotel_data = hotel_data[~hotel_data['house_rules'].str.contains('No parties')]
elif local_constraint['house rule'] == 'smoking':
hotel_data = hotel_data[~hotel_data['house_rules'].str.contains('No smoking')]
elif local_constraint['house rule'] == 'children under 10':
hotel_data = hotel_data[~hotel_data['house_rules'].str.contains('No children under 10')]
elif local_constraint['house rule'] == 'pets':
hotel_data = hotel_data[~hotel_data['house_rules'].str.contains('No pets')]
elif local_constraint['house rule'] == 'visitors':
hotel_data = hotel_data[~hotel_data['house_rules'].str.contains('No visitors')]
if days == 3:
if len(hotel_data) < 3:
raise ValueError("No hotel data available for the given constraints.")
elif days == 5:
if len(hotel_data) < 5:
raise ValueError("No hotel data available for the given constraints.")
elif days == 7:
if len(hotel_data) < 7:
raise ValueError("No hotel data available for the given constraints.")
if local_constraint['cuisine']:
# judge whether the cuisine is in the cuisine list
restaurant_data = restaurant_data[restaurant_data['Cuisines'].str.contains('|'.join(local_constraint['cuisine']))]
if days == 3:
if len(restaurant_data) < 3:
raise ValueError("No restaurant data available for the given constraints.")
elif days == 5:
if len(restaurant_data) < 5:
raise ValueError("No restaurant data available for the given constraints.")
elif days == 7:
if len(restaurant_data) < 7:
raise ValueError("No restaurant data available for the given constraints.")
# hotel_data = filter.run(hotel_data, local_constraint)
# restaurant_data = filter.run(restaurant_data, local_constraint)
# flight_data = filter.run(flight_data, local_constraint)
# Calculate budgets for all three modes
budgets = {}
for mode in ["lowest", "highest", "average"]:
if local_constraint and local_constraint['transportation'] == 'self driving':
flight_budget = eval(distanceMatrix.run(org, dest)['cost'].replace("$","")) * multipliers[days]["flight"]
else:
flight_budget = estimate_budget(flight_data["Price"].tolist(), mode) * multipliers[days]["flight"]
hotel_budget = estimate_budget(hotel_data["price"].tolist(), mode) * multipliers[days]["hotel"]
restaurant_budget = estimate_budget(restaurant_data["Average Cost"].tolist(), mode) * multipliers[days]["restaurant"]
total_budget = flight_budget + hotel_budget + restaurant_budget
budgets[mode] = total_budget
return budgets