Python pandas lambda 比较 Z6A8064B5DF479455500553C47DCZ 中的两列

Question

我在a，b中有两列货币；

两列中都有一些单元格为空。

我希望创建第三列 c 来识别以下逻辑：

if a == b then display 'same'
elif a == None then display 'a missing'
elif b == None then display 'b missing'
elif a == None and b == None then display 'all missing'
else 'diff currency'.

这些是我下面的代码。 它只是返回“相同”或“差异货币”，两者之间没有任何关系。

请在这里阐明我的语法或逻辑缺陷。 太感谢了！

import pandas as pd

# list of currencies
a = list(('USD USD CAD nan JMD nan HKD CAD').split())

b = list(('USD CAD RMB HKD nan nan USD EUR').split())


# df
df = pd.DataFrame(list(zip(a, b)), columns=['a', 'b'])

df = df.replace('nan', '')


df['c'] = df.apply(lambda x: 'Same' if x['a'] == x['b'] 
                   else ('a missing' if x['a']==None
                         else ('b missing' if x['b']==None 
                         else ('a & b missing' if x['a']==None and x['b']==None
                         else 'diff currency'))), axis=1)

Answer 1

如果您学习如何使用矢量化函数，那就更好了。 它们都是惯用的 pandas 并且速度极快。 使用np.select ：

a = df["a"]
b = df["b"]
df["c"] = np.select(
    [a.isna() & b.isna(), a.isna(), b.isna(), np.isclose(a, b)],
    ["all missing", "a missing", "b missing", "same"],
    "diff currency",
)

Answer 2

您可以为此使用np.select 。

import pandas as pd
import numpy as np

# list of currencies
a = list(('USD USD CAD nan JMD nan HKD CAD').split())

b = list(('USD CAD RMB HKD nan nan USD EUR').split())

# df
df = pd.DataFrame(list(zip(a, b)), columns=['a', 'b'])

# change string `nan` into actual NaN values
df = df.replace('nan', np.nan)

condlist = [df.a == df.b, df.isna().all(axis=1), df.a.isna(), df.b.isna()]
choicelist = ['same', 'all missing', 'a missing', 'b missing']

df['c'] = np.select(condlist,choicelist,default='diff currency')
print(df)

     a    b              c
0  USD  USD           same
1  USD  CAD  diff currency
2  CAD  RMB  diff currency
3  NaN  HKD      a missing
4  JMD  NaN      b missing
5  NaN  NaN    all missing
6  HKD  USD  diff currency
7  CAD  EUR  diff currency