# Connect to Google Drive
from google.colab import drive
drive.mount("/content/gdrive")

import csv

def load_airbnb_data(filename):
    """Load Airbnb data from CSV file into a list of dictionaries"""
    listings = []

    # Fields that should be converted to float
    numeric_fields = [
        'id', 'host_id', 'latitude', 'longitude', 'price',
        'minimum_nights', 'number_of_reviews', 'reviews_per_month',
        'calculated_host_listings_count', 'availability_365',
        'number_of_reviews_ltm'
    ]

    with open(filename, 'r', encoding='utf-8') as file:
        reader = csv.DictReader(file)
        for row in reader:
            # Try to convert numeric fields
            skip_row = False
            for field in numeric_fields:
                if field in row and row[field]:  # Check if field exists and is not empty
                    try:
                        row[field] = float(row[field])
                    except ValueError:
                        # Skip this entire row if any numeric conversion fails
                        skip_row = True
                        break
                else:
                    # Set empty numeric fields to 0.0
                    row[field] = 0.0

            if not skip_row:
                listings.append(row)
    return listings

# Load the data
listings = load_airbnb_data('/content/gdrive/MyDrive/datasets/NYC-Airbnb-2023.csv')
print(f"Loaded {len(listings)} listings")
print("First 5 listings:")
for listing in listings[:5]:
    print(listing)