Pandas多列GroupBy问题求助：添加Name列后运行异常或超时

问题背景 #

我是Pandas新手，写了一段分组统计的代码：

for x in data_cp.Owner_Type.unique():
 print(data_cp[(data_cp.Owner_Type == x)].groupby(['Owner_Type','Year','Transmission','Fuel_Type','New_Price'])
 ['Year'].count().sort_values(ascending=False).head(5))

但往分组列里加Name后，程序要么崩溃要么超时，根本没法正常分组。我现在需要完成两个统计需求：

1. 按Year、Owner_Type、Name、Fuel_Type统计年度汽车销量
2. 按城市（比如Kolkata、Delhi）结合年份统计对应汽车销量

数据样例如下：

data.head().to_dict() 

{'S.No.': {0: 0, 1: 1, 2: 2, 3: 3, 4: 4},
 'Name': {0: 'Maruti Wagon R LXI CNG',
  1: 'Hyundai Creta 1.6 CRDi SX Option',
  2: 'Honda Jazz V',
  3: 'Maruti Ertiga VDI',
  4: 'Audi A4 New 2.0 TDI Multitronic'},
 'Location': {0: 'Mumbai',
  1: 'Pune',
  2: 'Chennai',
  3: 'Chennai',
  4: 'Coimbatore'},
 'Year': {0: 2010, 1: 2015, 2: 2011, 3: 2012, 4: 2013},
 'Kilometers_Driven': {0: 72000, 1: 41000, 2: 46000, 3: 87000, 4: 40670},
 'Fuel_Type': {0: 'CNG', 1: 'Diesel', 2: 'Petrol', 3: 'Diesel', 4: 'Diesel'},
 'Transmission': {0: 'Manual',
  1: 'Manual',
  2: 'Manual',
  3: 'Manual',
  4: 'Automatic'},
 'Owner_Type': {0: 'First', 1: 'First', 2: 'First', 3: 'First', 4: 'Second'},
 'Mileage': {0: '26.6 km/kg',
  1: '19.67 kmpl',
  2: '18.2 kmpl',
  3: '20.77 kmpl',
  4: '15.2 kmpl'},
 'Engine': {0: '998 CC',
  1: '1582 CC',
  2: '1199 CC',
  3: '1248 CC',
  4: '1968 CC'},
 'Power': {0: '58.16 bhp',
  1: '126.2 bhp',
  2: '88.7 bhp',
  3: '88.76 bhp',
  4: '140.8 bhp'},
 'Seats': {0: 5.0, 1: 5.0, 2: 5.0, 3: 7.0, 4: 5.0},
 'New_Price': {0: 5.51, 1: 16.06, 2: 8.61, 3: 11.27, 4: 53.14},
 'Price': {0: 1.75, 1: 12.5, 2: 4.5, 3: 6.0, 4: 17.74}}

期望的分组效果参考（以2013年首次车主的Honda Amaze S i-Dtech为例）：

197,Honda Amaze S i-Dtech,Kolkata,2013,50000,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.25
346,Honda Amaze S i-Dtech,Kolkata,2013,57000,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.45
1662,Honda Amaze S i-Dtech,Delhi,2013,89429,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.65
1840,Honda Amaze S i-Dtech,Delhi,2013,90000,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.8
2953,Honda Amaze S i-Dtech,Kolkata,2013,38755,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.2
3438,Honda Amaze S i-Dtech,Kolkata,2013,39575,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.25
4740,Honda Amaze S i-Dtech,Bangalore,2013,61510,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,4.75
5403,Honda Amaze S i-Dtech,Kolkata,2013,46992,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,3.75
6626,Honda Amaze S i-Dtech,Kolkata,2013,51637,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,
6692,Honda Amaze S i-Dtech,Chennai,2013,75000,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,
7237,Honda Amaze S i-Dtech,Kolkata,2013,38568,Diesel,Manual,First,25.8 kmpl,1498 CC,98.6 bhp,5,8.54,

解决方案 #

问题根源 #

原代码绕了个弯路：先循环遍历Owner_Type的每个唯一值，再分别分组统计。这种操作完全冗余，当加入Name后，分组的组合数暴增，再加上循环带来的额外开销，直接导致超时或崩溃。正确的做法是直接对整个数据集做一次分组，不需要循环。

1. 按Year、Owner_Type、Name、Fuel_Type统计销量 #

# 用size()统计分组行数，比count()更高效（无需处理NaN）
sales_by_car = data_cp.groupby(['Year', 'Owner_Type', 'Name', 'Fuel_Type'], as_index=False).size()
# 把统计列名改成更直观的“销量”
sales_by_car.rename(columns={'size': '销量'}, inplace=True)
# 按年份升序、销量降序排序，方便查看每年的热门车型
sales_by_car.sort_values(['Year', '销量'], ascending=[True, False], inplace=True)
# 查看前10条结果
print(sales_by_car.head(10))

2. 按城市+年度统计销量 #

# 按城市和年份分组统计
sales_by_city_year = data_cp.groupby(['Location', 'Year'], as_index=False).size()
sales_by_city_year.rename(columns={'size': '销量'}, inplace=True)
# 按城市名称和年份排序
sales_by_city_year.sort_values(['Location', 'Year'], ascending=True, inplace=True)
# 查看前10条结果
print(sales_by_city_year.head(10))

3. 查看特定车型的详细分组数据 #

如果要像示例那样，查看某款车的所有原始记录，不需要用groupby，直接筛选即可：

# 筛选2013年、首次车主的Honda Amaze S i-Dtech
target_data = data_cp[(data_cp['Name'] == 'Honda Amaze S i-Dtech') & 
                      (data_cp['Year'] == 2013) & 
                      (data_cp['Owner_Type'] == 'First')]
# 按城市排序后输出
target_data.sort_values('Location', inplace=True)
print(target_data)

额外优化建议 #

如果数据集特别大，可以先只保留需要的列再分组，减少内存占用：

# 只保留分组和统计必需的列
reduced_data = data_cp[['Year', 'Owner_Type', 'Name', 'Fuel_Type', 'Location']]
# 再执行分组统计
sales_by_car = reduced_data.groupby(['Year', 'Owner_Type', 'Name', 'Fuel_Type']).size().reset_index(name='销量')

内容的提问来源于stack exchange，提问作者panda_newbie