utils2.py

from collections import Counter
import pandas as pd
import numpy as np
from scipy.spatial import cKDTree


df_amenities = pd.read_csv("df_indonesia.csv").rename(
    columns={"latitude":"lat", "longitude":"lon"}
)
df_banks = pd.read_csv("df_bank_indonesia.csv").rename(
    columns={"latitude":"lat", "longitude":"lon"}
)

df_amenities["fsq_category_labels"] = df_amenities["fsq_category_labels"].apply(
    lambda x: eval(x)
)

bank_coords = df_banks[['lat','lon']].values
tree_banks = cKDTree(bank_coords)

amenity_coords = df_amenities[['lat','lon']].values
tree_amenities = cKDTree(amenity_coords)

# DATASET_COLUMNS = [
#     'Dining and Drinking', 'Community and Government', 'Retail',
#        'Business and Professional Services', 'Landmarks and Outdoors',
#        'Arts and Entertainment', 'Health and Medicine',
#        'Travel and Transportation', 'Sports and Recreation',
#        'Event'
# ]

DATASET_COLUMNS = [
    "Community and Government > Education",
    "Landmarks and Outdoors > Structure",
    "Arts and Entertainment > Art Gallery",
    "Arts and Entertainment > Movie Theater",
    "Retail > Department Store",
    "Business and Professional Services > Office",
    "Travel and Transportation > Road",
    "Dining and Drinking > Restaurant",
    "Community and Government > Residential Building",
    "Dining and Drinking > Breakfast Spot",
    "Business and Professional Services",
    "Arts and Entertainment > Amusement Park",
    "Travel and Transportation > Lodging",
    "Arts and Entertainment",
    "Dining and Drinking > Food Truck",
    "Dining and Drinking > Bar",
    "Retail > Sporting Goods Retail",
    "Retail > Computers and Electronics Retail",
    "Arts and Entertainment > Arcade",
    "Dining and Drinking > Cafe, Coffee, and Tea House",
    "Travel and Transportation > Transport Hub",
    "Health and Medicine > Physician",
    "Community and Government > Government Building",
    "Business and Professional Services > Convention Center",
    "Arts and Entertainment > Performing Arts Venue",
    "Landmarks and Outdoors > Field",
    "Business and Professional Services > Financial Service",
    "Landmarks and Outdoors > Park",
    "Sports and Recreation > Water Sports",
    "Landmarks and Outdoors > Other Great Outdoors",
    "Event > Convention",
    "Retail > Shopping Mall",
    "Business and Professional Services > Distribution Center",
    "Business and Professional Services > Automotive Service",
    "Health and Medicine > Hospital",
    "Dining and Drinking > Snack Place",
    "Business and Professional Services > Event Space",
    "Dining and Drinking > Food Court",
    "Travel and Transportation > Bike Rental",
    "Travel and Transportation > Parking",
    "Business and Professional Services > Radio Station",
    "Health and Medicine > Dentist",
    "Landmarks and Outdoors > Beach",
    "Retail > Flea Market",
    "Retail > Fashion Retail",
    "Retail > Food and Beverage Retail",
    "Retail > Office Supply Store",
    "Community and Government > Spiritual Center",
    "Health and Medicine > Medical Center",
    "Dining and Drinking > Bakery",
    "Dining and Drinking > Cafeteria",
    "Retail > Convenience Store",
    "Arts and Entertainment > Public Art",
    "Retail > Newsstand",
    "Retail > Furniture and Home Store",
    "Business and Professional Services > Auditorium",
    "Landmarks and Outdoors > Garden",
    "Community and Government > Library",
    "Community and Government > Organization",
    "Business and Professional Services > Health and Beauty Service",
    "Dining and Drinking > Food Stand",
    "Retail > Pharmacy",
    "Retail > Record Store",
    "Arts and Entertainment > Night Club",
    "Landmarks and Outdoors > Farm",
    "Community and Government > Social Club",
    "Sports and Recreation > Gym and Studio",
    "Sports and Recreation > Racquet Sports",
    "Retail > Costume Store",
    "Sports and Recreation > Soccer",
    "Event > Entertainment Event",
    "Retail > Market",
    "Sports and Recreation",
    "Community and Government > Town Hall",
    "Arts and Entertainment > Water Park",
    "Landmarks and Outdoors",
    "Business and Professional Services > Shoe Repair Service",
    "Retail > Automotive Retail",
    "Dining and Drinking > Dessert Shop",
    "Retail > Music Store",
    "Business and Professional Services > Factory",
    "Community and Government > Assisted Living",
    "Travel and Transportation > Transportation Service",
    "Health and Medicine > Medical Lab",
    "Retail",
    "Travel and Transportation > Rest Area",
    "Landmarks and Outdoors > Hiking Trail",
    "Sports and Recreation > Martial Arts Dojo",
    "Business and Professional Services > Laundry Service",
    "Retail > Pet Supplies Store",
    "Business and Professional Services > Design Studio",
    "Community and Government > Housing Development",
    "Business and Professional Services > Construction",
    "Business and Professional Services > Real Estate Service",
    "Landmarks and Outdoors > Historic and Protected Site",
    "Business and Professional Services > Photography Service",
    "Landmarks and Outdoors > Plaza",
    "Business and Professional Services > Child Care Service",
    "Event > Other Event",
    "Arts and Entertainment > Comedy Club",
    "Arts and Entertainment > Strip Club",
    "Arts and Entertainment > Casino",
    "Business and Professional Services > Legal Service",
    "Retail > Miscellaneous Store",
    "Travel and Transportation",
    "Retail > Boutique",
    "Sports and Recreation > Race Track",
    "Event > Marketplace",
    "Sports and Recreation > Baseball",
    "Retail > Arts and Crafts Store",
    "Travel and Transportation > Moving Target",
    "Community and Government > Cemetery",
    "Business and Professional Services > Wholesaler",
    "Business and Professional Services > Advertising Agency",
    "Arts and Entertainment > Stadium",
    "Arts and Entertainment > Country Dance Club",
    "Landmarks and Outdoors > Bridge",
    "Health and Medicine > Alternative Medicine Clinic",
    "Arts and Entertainment > Zoo",
    "Business and Professional Services > Food and Beverage Service",
    "Business and Professional Services > Storage Facility",
    "Business and Professional Services > Recycling Facility",
    "Business and Professional Services > Funeral Home",
    "Arts and Entertainment > Bowling Alley",
    "Retail > Hardware Store",
    "Retail > Stationery Store",
    "Retail > Bookstore",
    "Retail > Smoke Shop",
    "Sports and Recreation > Running and Track",
    "Retail > Vape Store",
    "Landmarks and Outdoors > Campground",
    "Retail > Gift Store",
    "Landmarks and Outdoors > Palace",
    "Business and Professional Services > Recording Studio",
    "Landmarks and Outdoors > States and Municipalities",
    "Community and Government > Community Center",
    "Travel and Transportation > Fuel Station",
    "Retail > Cosmetics Store",
    "Retail > Hobby Store",
    "Travel and Transportation > Travel Lounge",
    "Community and Government > Polling Place",
    "Retail > Baby Store",
    "Business and Professional Services > Telecommunication Service",
    "Health and Medicine > Healthcare Clinic",
    "Dining and Drinking > Donut Shop",
    "Travel and Transportation > Tourist Information and Service",
    "Dining and Drinking > Juice Bar",
    "Retail > Antique Store",
    "Retail > Toy Store",
    "Business and Professional Services > Event Service",
    "Landmarks and Outdoors > Monument",
    "Sports and Recreation > Snow Sports",
    "Travel and Transportation > Travel Agency",
    "Landmarks and Outdoors > Stable",
    "Arts and Entertainment > Museum",
    "Business and Professional Services > Home Improvement Service",
    "Retail > Flower Store",
    "Business and Professional Services > Technology Business",
    "Travel and Transportation > Electric Vehicle Charging Station",
    "Business and Professional Services > Business Center",
    "Landmarks and Outdoors > Lake",
    "Business and Professional Services > Film Studio",
    "Dining and Drinking > Vineyard",
    "Dining and Drinking > Distillery",
    "Business and Professional Services > Warehouse",
    "Business and Professional Services > Pet Service",
    "Dining and Drinking > Brewery",
    "Retail > Eyecare Store",
    "Health and Medicine",
    "Business and Professional Services > Tailor",
    "Retail > Board Store",
    "Sports and Recreation > Golf",
    "Business and Professional Services > Audiovisual Service",
    "Business and Professional Services > Manufacturer",
    "Retail > Shopping Plaza",
    "Business and Professional Services > Publisher",
    "Retail > Print Store",
    "Sports and Recreation > Gymnastics",
    "Landmarks and Outdoors > Mountain",
    "Retail > Perfume Store",
    "Arts and Entertainment > Gaming Cafe",
    "Travel and Transportation > Boat or Ferry",
    "Community and Government > Cultural Center",
    "Business and Professional Services > Employment Agency",
    "Arts and Entertainment > Psychic and Astrologer",
    "Retail > Big Box Store",
    "Arts and Entertainment > Pool Hall",
    "Community and Government > Animal Shelter",
    "Landmarks and Outdoors > Surf Spot",
    "Landmarks and Outdoors > Castle",
    "Arts and Entertainment > Internet Cafe",
    "Health and Medicine > Emergency Service",
    "Sports and Recreation > Volleyball Court",
    "Business and Professional Services > Security and Safety",
    "Sports and Recreation > Basketball",
    "Business and Professional Services > Shipping, Freight, and Material Transportation Service",
    "Health and Medicine > Veterinarian",
    "Health and Medicine > Acupuncture Clinic",
    "Dining and Drinking > Creperie",
    "Retail > Outlet Store",
    "Dining and Drinking > Bagel Shop",
    "Landmarks and Outdoors > Scenic Lookout",
    "Dining and Drinking > Winery",
    "Retail > Garden Center",
    "Business and Professional Services > Import and Export Service",
    "Sports and Recreation > Skating",
    "Health and Medicine > Physical Therapy Clinic",
    "Landmarks and Outdoors > Botanical Garden",
    "Travel and Transportation > Train",
    "Retail > Warehouse or Wholesale Store",
    "Retail > Luggage Store",
    "Business and Professional Services > TV Station",
    "Retail > Pop-Up Store",
    "Business and Professional Services > Entertainment Service",
    "Landmarks and Outdoors > Sculpture Garden",
    "Landmarks and Outdoors > Harbor or Marina",
    "Health and Medicine > Home Health Care Service",
    "Retail > Adult Store",
    "Health and Medicine > Chiropractor",
    "Health and Medicine > Mental Health Service",
    "Business and Professional Services > Art Studio",
    "Business and Professional Services > Metals Supplier",
    "Landmarks and Outdoors > Island",
    "Travel and Transportation > RV Park",
    "Landmarks and Outdoors > Lighthouse",
    "Retail > Construction Supplies Store",
    "Business and Professional Services > Plastics Supplier",
    "Retail > Video Store",
    "Event > Conference",
    "Retail > Tobacco Store",
    "Business and Professional Services > Agriculture and Forestry Service",
    "Landmarks and Outdoors > River",
    "Business and Professional Services > Chemicals and Gasses Manufacturer",
    "Business and Professional Services > Repair Service",
    "Business and Professional Services > Rubber Supplier",
    "Landmarks and Outdoors > Roof Deck",
    "Retail > Vintage and Thrift Store",
    "Sports and Recreation > Fishing Area",
    "Sports and Recreation > Sports Club",
    "Business and Professional Services > Business Service",
    "Business and Professional Services > Media Agency",
    "Landmarks and Outdoors > Hot Spring",
    "Landmarks and Outdoors > Well",
    "Business and Professional Services > Insurance Agency",
    "Community and Government > Summer Camp",
    "Landmarks and Outdoors > Bay",
    "Business and Professional Services > Industrial Equipment Supplier",
    "Retail > Comic Book Store",
    "Travel and Transportation > Toll Booth",
    "Dining and Drinking",
    "Arts and Entertainment > Fair",
    "Retail > Souvenir Store",
    "Sports and Recreation > Paintball Field",
    "Retail > Drugstore",
    "Arts and Entertainment > Exhibit",
    "Retail > Supplement Store",
    "Retail > Outdoor Supply Store",
    "Arts and Entertainment > Escape Room",
    "Health and Medicine > Optometrist",
    "Business and Professional Services > Engineer",
    "Sports and Recreation > Indoor Play Area",
    "Retail > Leather Goods Store",
    "Community and Government > Public Bathroom",
    "Business and Professional Services > Electrical Equipment Supplier",
    "Travel and Transportation > Baggage Locker",
    "Arts and Entertainment > Go Kart Track",
    "Arts and Entertainment > Circus",
    "Sports and Recreation > Hockey",
    "Landmarks and Outdoors > Forest",
    "Business and Professional Services > Computer Repair Service",
    "Business and Professional Services > Waste Management Service",
    "Travel and Transportation > Platform",
    "Retail > Textiles Store",
    "Business and Professional Services > Scientific Equipment Supplier",
    "Landmarks and Outdoors > Pedestrian Plaza",
    "Community and Government > Utility Company",
    "Community and Government > Public and Social Service",
    "Business and Professional Services > Petroleum Supplier",
    "Business and Professional Services > Wedding Hall",
    "Landmarks and Outdoors > Nature Preserve",
    "Business and Professional Services > Ballroom",
    "Community and Government > Prison",
    "Sports and Recreation > Recreation Center",
    "Sports and Recreation > Gun Range",
    "Dining and Drinking > Smoothie Shop",
    "Dining and Drinking > Night Market",
    "Retail > Discount Store",
    "Business and Professional Services > Welding Service",
    "Travel and Transportation > Pier",
    "Landmarks and Outdoors > Bathing Area",
    "Business and Professional Services > Print, TV, Radio and Outdoor Advertising Service",
    "Business and Professional Services > Online Advertising Service",
    "Arts and Entertainment > Aquarium",
    "Arts and Entertainment > Roller Rink",
    "Community and Government > Trailer Park",
    "Business and Professional Services > Paper Supplier",
    "Retail > Framing Store",
    "Landmarks and Outdoors > Tunnel",
    "Health and Medicine > Urgent Care Center",
    "Community and Government > Rehabilitation Center",
    "Landmarks and Outdoors > Fountain",
    "Arts and Entertainment > Planetarium",
    "Sports and Recreation > Cricket Ground",
    "Landmarks and Outdoors > Volcano",
    "Business and Professional Services > Research Laboratory",
    "Business and Professional Services > Equipment Rental Service",
    "Community and Government",
    "Retail > Medical Supply Store",
    "Landmarks and Outdoors > Bike Trail",
    "Business and Professional Services > Outdoor Event Space",
    "Sports and Recreation > Rugby",
    "Business and Professional Services > Laboratory",
    "Business and Professional Services > Water Treatment Service",
    "Business and Professional Services > Entertainment Agency",
    "Retail > Pawn Shop",
    "Arts and Entertainment > Salsa Club",
    "Landmarks and Outdoors > Tree",
    "Travel and Transportation > Toll Plaza",
    "Travel and Transportation > Port",
    "Landmarks and Outdoors > Rock Climbing Spot",
    "Business and Professional Services > Creative Service",
    "Business and Professional Services > Research Station",
    "Business and Professional Services > Refrigeration and Ice Supplier",
    "Business and Professional Services > Rental Service",
    "Travel and Transportation > Border Crossing",
    "Business and Professional Services > Industrial Estate",
    "Business and Professional Services > Tutoring Service",
    "Business and Professional Services > Laundromat",
    "Arts and Entertainment > Mini Golf Course",
    "Retail > Packaging Supply Store",
    "Business and Professional Services > Translation Service",
    "Retail > Duty-free Store",
    "Community and Government > Observatory",
    "Retail > Knitting Store",
    "Travel and Transportation > Cruise",
    "Business and Professional Services > Management Consultant",
    "Retail > Betting Shop",
    "Retail > Outlet Mall",
    "Retail > Auction House",
    "Travel and Transportation > Cable Car",
    "Business and Professional Services > Power Plant",
    "Landmarks and Outdoors > Dive Spot",
    "Health and Medicine > Maternity Clinic",
    "Health and Medicine > Women's Health Clinic",
    "Health and Medicine > Nutritionist",
    "Retail > Mobility Store",
    "Business and Professional Services > Renewable Energy Service",
    "Landmarks and Outdoors > Hill",
    "Health and Medicine > Hospice",
    "Business and Professional Services > Machine Shop",
    "Landmarks and Outdoors > Memorial Site",
    "Landmarks and Outdoors > Cave",
    "Travel and Transportation > Truck Stop",
    "Business and Professional Services > Logging Service",
    "Landmarks and Outdoors > Waterfront",
    "Sports and Recreation > Skydiving Center",
    "Arts and Entertainment > Disc Golf",
    "Business and Professional Services > Promotional Item Service",
    "Business and Professional Services > Lottery Retailer",
]

def compute_features(candidate_point, radius=0.005):
    lat, lon = candidate_point

    # Banks
    bank_idxs = tree_banks.query_ball_point([lat, lon], r=radius)

    print("[BANK]", bank_idxs)
    
    n_banks = len(bank_idxs)
    if n_banks > 0:
        neighbors = df_banks.iloc[bank_idxs]
        mean_dist_banks = np.mean(np.sqrt((neighbors['lat']-lat)**2 + (neighbors['lon']-lon)**2))
        min_dist_bank = np.min(np.sqrt((neighbors['lat']-lat)**2 + (neighbors['lon']-lon)**2))
    else:
        mean_dist_banks = radius
        min_dist_bank = radius

    # Amenities
    amenity_idxs = tree_amenities.query_ball_point([lat, lon], r=radius)
    amenities = df_amenities.iloc[amenity_idxs]

    total_amenities = len(amenities)

    # Flatten all category IDs
    # for cats in amenities['fsq_category_labels']:
    all_category_ids = [">".join(cats[0].split(">")[:2]).strip() for cats in amenities['fsq_category_labels'] if len(cats)>0]
    category_diversity = len(set(all_category_ids))

    features = {
        'num_banks_in_radius': n_banks,
        # 'mean_dist_banks': mean_dist_banks,
        # 'min_dist_bank': min_dist_bank,
        'total_amenities': total_amenities,
        'category_diversity': category_diversity
    }

    # Count occurrences per category
    print("[CATEGORIES]", all_category_ids)
    count_per_category = Counter(all_category_ids)
    for feat in DATASET_COLUMNS:
        print("[FEAT]",feat)
    # for cat, cnt in count_per_category.items():
        features[f'num_{feat}'] = count_per_category.get(feat, 0)

    
    # # Count occurrences of first category
    # first_categories = [cats[0] for cats in amenities['fsq_category_ids'] if len(cats)>0]
    # count_first_category = Counter(first_categories)
    # for cat, cnt in count_first_category.items():
    #     features[f'num_first_{cat}'] = cnt

    return features