分割数据帧的两列时出现值错误。（.all（）和.any（）也不起作用）

Question

这是数据帧信息：

new_final.info() 

<class 'pandas.core.frame.DataFrame'>
Int64Index: 12 entries, 1 to 13
Data columns (total 9 columns):
 #   Column           Non-Null Count  Dtype 
---  ------           --------------  ----- 
 0   DorW             12 non-null     object
 1   date             12 non-null     object
 2   source           12 non-null     object
 3   AU               12 non-null     int32 
 4   impressions      12 non-null     int32 
 5   clicks           12 non-null     int32 
 6   pdp              12 non-null     int32 
 7   CABN             12 non-null     int32 
 8   order_completed  12 non-null     int32 
dtypes: int32(6), object(3)
memory usage: 672.0+ bytes

我想要一个新的专栏，它是pdp/impression的一部分，我已经尝试了所有这些方法，但仍然得到这些错误，我该如何修复它

new_final['impression_to_pdp'] = (new_final['pdp']/new_final['impressions']) 
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

new_final['impression_to_pdp'] = (new_final['pdp']/new_final['impressions']).any() 
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

new_final['impression_to_pdp'] = (new_final['pdp']/new_final['impressions']).all() 
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

数据帧

Answer 1

纯粹的尝试和错误，您是否在没有括号的情况下尝试过

new_final['impression_to_pdp'] = new_final['pdp']/new_final['impressions']

Answer 2

如果它们是包含在列表中的普通整数，则应该以这种方式工作

new_final['impression_to_pdp'] = new_final['pdp']/new_final['impressions']

但是，如果它们是numpy数组，那么您必须像

import numpy as np
new_final['impression_to_pdp'] = np.divide(new_final['pdp'].values, new_final['impressions'].values)