hotel_recommendations1.py

import pandas as pd
from sklearn.metrics import euclidean_distances
from sklearn.preprocessing import StandardScaler


def get_hotel_recommendations(df: pd.DataFrame, anchor_id: int) -> pd.DataFrame:
    # features used to compute the similarity
    features = ['lat', 'lng', 'avg_rate', 'star_rating', 'user_rating']
    
    # create the features - make the anchor be the first row in the dataframe
    df_sorted = df.copy()
    df_sorted = pd.concat([df_sorted[df_sorted['hotel_id'] == anchor_id],
                           df_sorted[df_sorted['hotel_id'] != anchor_id]])
    df_features = df_sorted[features].copy()
    df_features = normalize_features(df_features)
    
    # compute the distances
    X = df_features.values
    Y = df_features.values[0].reshape(1, -1)
    distances = euclidean_distances(X, Y)
    
    df_sorted['similarity_distance'] = distances
    return df_sorted.sort_values('similarity_distance').reset_index(drop=True)


def normalize_features(df: pd.DataFrame) -> pd.DataFrame:
    df_norm = df.copy()
    for col in df_norm.columns:
        # fill any NaN's with the mean
        df_norm[col] = df_norm[col].fillna(df_norm[col].mean())
        df_norm[col] = StandardScaler().fit_transform(df_norm[col].values.reshape(-1, 1))
    return df_norm