hotel_recommendations2.py

from typing import Dict

import pandas as pd
from sklearn.metrics import euclidean_distances


def get_personalized_hotel_recommendations(df: pd.DataFrame, user_features: Dict) -> pd.DataFrame:
    # features used to compute the similarity
    features = ['distance', 'avg_rate', 'star_rating', 'user_rating']
    
    # create the features
    df_features = df[features].copy()
    df_features = normalize_features(df_features)
    
    # artifically set the distance of the user anchor to be the min of the
    # normalized distance
    user_features['distance'] = df_features['distance'].min()
    df_user = pd.DataFrame([user_features])
    df_features = pd.concat([df_user, df_features], sort=False)
    
    # compute the distances
    X = df_features.values
    Y = df_features.values[0].reshape(1, -1)
    distances = euclidean_distances(X, Y)
    
    df_sorted = df.copy()
    df_sorted['similarity_distance'] = distances[1:]
    return df_sorted.sort_values('similarity_distance').reset_index(drop=True)