Numpy求平均触发ufunc 'divide'错误，求问题排查与解决

问题根源 #

你当前的代码存在逻辑错误：

• 你创建的fuel_typesx、fuel_typesz等数组，存储的是燃料类型的字符串值（比如重复的'X'、'Z'），而np.average()只能对数值类型数据计算平均值，字符串无法进行除法运算，因此触发了类型不兼容的报错。
• 你的需求是计算不同燃料类型对应的平均油耗，但你错误地对燃料类型本身求平均，而非对应燃料类型的油耗数据。

修正后的代码 #

通过布尔索引筛选对应燃料类型的油耗数据，再计算平均值：

import numpy as np
import pandas as pd

cars = pd.read_csv('data/co2_emissions_canada.csv')
cars_fuel_types = cars['Fuel Type'].to_numpy()
cars_fuel_consumption = cars['Fuel Consumption Comb (L/100 km)'].to_numpy()

# 计算普通汽油（X）的平均油耗
fuel_x_consumption = cars_fuel_consumption[cars_fuel_types == 'X']
print(np.average(fuel_x_consumption))

# 计算 premium gasoline（Z）的平均油耗
fuel_z_consumption = cars_fuel_consumption[cars_fuel_types == 'Z']
print(np.average(fuel_z_consumption))

# 计算乙醇（E）的平均油耗
fuel_e_consumption = cars_fuel_consumption[cars_fuel_types == 'E']
print(np.average(fuel_e_consumption))

# 计算柴油（D）的平均油耗
fuel_d_consumption = cars_fuel_consumption[cars_fuel_types == 'D']
print(np.average(fuel_d_consumption))

更简洁的Pandas分组实现 #

既然已经使用Pandas，直接用分组功能可以更高效完成需求，无需手动处理Numpy数组：

import pandas as pd

cars = pd.read_csv('data/co2_emissions_canada.csv')
# 按燃料类型分组，计算每组的平均油耗
fuel_avg_result = cars.groupby('Fuel Type')['Fuel Consumption Comb (L/100 km)'].mean()
print(fuel_avg_result)

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