如何创建非空DataFrame？酒店推荐算法空DataFrame问题排查

问题描述 #

我在用一套酒店推荐算法，文本输出符合预期，但把推荐酒店列表和其他酒店列表转换为对应DataFrame时，始终得到空DataFrame。需要将这两类酒店数据分别存入独立的DataFrame，请求协助排查问题。

代码实现 #

import pandas as pd


def recommend_hotels(data, weights):

  df = pd.DataFrame(data)

  # Ask the user for the passenger count
  passenger_count = int(input("Enter the number of passengers: "))

  # Calculate the minimum number of rooms needed
  min_rooms_needed = passenger_count

  # Calculate the difference between total room availability and the passenger count
  df['room_diff'] = df['total_rooms'] - min_rooms_needed

  # Filter hotels with higher availability than the passenger count and sort by room availability in ascending order
   df = df[df['room_diff'] >= 0].sort_values(by='total_rooms')

  # If there are no hotels with higher availability, sort by room availability in ascending order
  if df.empty:
    df = data.copy()
    df = df.sort_values(by='total_rooms')

  # Initialize variables to track room assignments
  assigned_rooms = []
  remaining_passengers = passenger_count

  # Assign double rooms first
  for index, row in df.iterrows():
    double_rooms_available = row['double_rooms']
    if double_rooms_available > 0:
        assigned_double_rooms = min(double_rooms_available, remaining_passengers // 2)
        remaining_passengers -= assigned_double_rooms * 2
        assigned_rooms.append((row['hotel_name'], assigned_double_rooms, 'double'))

  # Assign single rooms if needed
   for index, row in df.iterrows():
    if remaining_passengers > 0:
        single_rooms_available = row['single_rooms']
        assigned_single_rooms = min(single_rooms_available, remaining_passengers)
        remaining_passengers -= assigned_single_rooms
        assigned_rooms.append((row['hotel_name'], assigned_single_rooms, 'single'))

  # Recommend hotels based on room assignments
    recommended_hotels = [hotel[0] for hotel in assigned_rooms]

  # Calculate the total number of rooms to book for all passengers
  total_rooms_to_book = min_rooms_needed

  # Create a DataFrame for recommended hotels
   recommended_hotels_df = 
 df[df['hotel_name'].isin(recommended_hotels)].sort_values(by='total_rooms', ascending=False)
recommended_hotels_df
# Print the initial recommendations
print(f"Recommended Hotels: {', '.join(recommended_hotels)}")
print(f"Book {total_rooms_to_book} rooms for {passenger_count} passengers.")

# List unrecommended hotels in descending order of room availability
unrecommended_hotels = set(data['hotel_name']) - set(recommended_hotels)

if unrecommended_hotels:
    unrecommended_df = df[df['hotel_name'].isin(unrecommended_hotels)].sort_values(by='total_rooms', ascending=False)
    #print(f"\nUnrecommended Hotels (Not in Recommended List) in Descending Order of Availability:")
    for index, row in unrecommended_df.iterrows():
        print(f"{row['hotel_name']} - Total Rooms: {row['total_rooms']}")

# List all other hotels that are not in the recommended list
other_hotels = set(data['hotel_name']) - set(recommended_hotels)

print(f"\nAll Other Hotels (Not in Recommended List): {', '.join(other_hotels)}")

# Create a DataFrame for all other hotels that are not in the recommended list
other_hotels_df = df[df['hotel_name'].isin(other_hotels)]

# Define your data and weights
 data = {
'hotel_name': ['A', 'B', 'C', 'D', 'E'],
'single_rooms': [10, 25, 8, 12, 7],
'double_rooms': [35, 18, 44, 110, 6],
'total_rooms': [45, 43, 52, 122, 13],
 }

 weights = {
 'num_hotels': 0.35,
 'transport': 0.3,
 'room_price': 0.15,
 'meals': 0.1,
 'availability': 0.075,
 'rooms': 0.025,
}
other_hotels_df


# Call the recommend_hotels function with your data and weights
recommend_hotels(data, weights)

当前输出 #

输出显示空DataFrame，无有效数据。

尝试的代码 #

other_hotels = set(data['hotel_name']) - set(recommended_hotels)
other_hotels_df =df[df['hotel_name'].isin(other_hotels)].sort_values(by='total_rooms', ascending=False)

期望输出 #

Recommended_df: #

UnRecommended_df: #

问题根源与修复方案 #

1. 核心问题 #

• 缩进错误导致变量作用域异常：recommended_hotels的定义被错误缩进，属于第二个for循环的内部代码，仅在循环满足条件时才会赋值，导致变量可能未定义或为空。
• DataFrame赋值语法错误：recommended_hotels_df的定义存在换行错误，无法正确初始化。
• 过滤后的数据丢失问题：当通过df = df[df['room_diff'] >= 0]过滤酒店后，部分酒店（如E）被排除在当前df之外，后续用该df筛选非推荐酒店时无法找到对应数据。
• 无返回值：函数内创建的DataFrame未返回，外部无法获取结果。

2. 修复后的完整代码 #

import pandas as pd


def recommend_hotels(data, weights):
    # 保存原始数据的DataFrame，避免过滤后丢失数据
    df_raw = pd.DataFrame(data)
    df = df_raw.copy()

    # 获取乘客数量
    passenger_count = int(input("Enter the number of passengers: "))
    min_rooms_needed = passenger_count

    # 计算房间余量
    df['room_diff'] = df['total_rooms'] - min_rooms_needed

    # 过滤可用房间足够的酒店并排序
    df = df[df['room_diff'] >= 0].sort_values(by='total_rooms')

    # 无满足条件的酒店时，使用原始数据排序
    if df.empty:
        df = df_raw.copy().sort_values(by='total_rooms')

    assigned_rooms = []
    remaining_passengers = passenger_count

    # 优先分配双人间
    for index, row in df.iterrows():
        double_rooms_available = row['double_rooms']
        if double_rooms_available > 0:
            assigned_double_rooms = min(double_rooms_available, remaining_passengers // 2)
            remaining_passengers -= assigned_double_rooms * 2
            assigned_rooms.append((row['hotel_name'], assigned_double_rooms, 'double'))

    # 分配单人间补全
    for index, row in df.iterrows():
        if remaining_passengers <= 0:
            break
        single_rooms_available = row['single_rooms']
        assigned_single_rooms = min(single_rooms_available, remaining_passengers)
        remaining_passengers -= assigned_single_rooms
        assigned_rooms.append((row['hotel_name'], assigned_single_rooms, 'single'))

    # 提取推荐酒店列表（修正缩进，确保全局作用域）
    recommended_hotels = [hotel[0] for hotel in assigned_rooms]
    total_rooms_to_book = min_rooms_needed

    # 基于原始DataFrame创建推荐/非推荐DataFrame
    recommended_hotels_df = df_raw[df_raw['hotel_name'].isin(recommended_hotels)].sort_values(by='total_rooms', ascending=False)
    unrecommended_hotels = set(data['hotel_name']) - set(recommended_hotels)
    unrecommended_df = df_raw[df_raw['hotel_name'].isin(unrecommended_hotels)].sort_values(by='total_rooms', ascending=False)

    # 打印输出
    print(f"Recommended Hotels: {', '.join(recommended_hotels)}")
    print(f"Book {total_rooms_to_book} rooms for {passenger_count} passengers.")
    
    if unrecommended_hotels:
        print(f"\nUnrecommended Hotels (Not in Recommended List) in Descending Order of Availability:")
        for index, row in unrecommended_df.iterrows():
            print(f"{row['hotel_name']} - Total Rooms: {row['total_rooms']}")
    
    print(f"\nAll Other Hotels (Not in Recommended List): {', '.join(unrecommended_hotels)}")

    # 返回两个DataFrame
    return recommended_hotels_df, unrecommended_df


# 定义数据与权重
data = {
    'hotel_name': ['A', 'B', 'C', 'D', 'E'],
    'single_rooms': [10, 25, 8, 12, 7],
    'double_rooms': [35, 18, 44, 110, 6],
    'total_rooms': [45, 43, 52, 122, 13],
}

weights = {
    'num_hotels': 0.35,
    'transport': 0.3,
    'room_price': 0.15,
    'meals': 0.1,
    'availability': 0.075,
    'rooms': 0.025,
}

# 调用函数并获取结果
recommended_df, unrecommended_df = recommend_hotels(data, weights)
print("\nRecommended_df:")
print(recommended_df)
print("\nUnRecommended_df:")
print(unrecommended_df)

3. 修复要点说明 #

• 新增df_raw保存原始数据，避免过滤后丢失非推荐酒店的信息。
• 修正recommended_hotels的缩进，确保变量在全局作用域内正确赋值。
• 修复recommended_hotels_df的语法错误，基于原始DataFrame筛选数据。
• 函数返回两个DataFrame，外部可以直接获取结果。
• 在单人间分配循环中增加break条件，优化性能。

内容的提问来源于stack exchange，提问作者Noob Coder