不要将 lambda 函数应用于缺失值

Question

我有一个列中包含患者诊断的数据框，并使用 pandas 我想对诊断进行二分 ==> ISM，非 ISM。我试过这个

df["initial_diagnosis"] = df["initial_diagnosis"].apply(lambda x: x if x=="ISM" else "non ISM")

但它也将“非 ISM”分配给缺失值。有没有办法做同样的事情并保持缺失值不变？

我尝试编码的专栏如下所示：

initial_diagnosis    I
ISM 
ISM
WDSM
NaN
ISM
SSM
CM
ASM
ISM

Answer 1

我认为它应该有效。可能缺失值是空字符串或者只是None，我只能猜测

missing_values = {...} # Set of values you want to keep
df["initial_diagnosis"] = df["initial_diagnosis"].apply(lambda x: x if x=="ISM" or x in missing_values else "non ISM")

编辑:

import pandas as pd
from numpy import nan

data = pd.read_csv("test.csv")
print(data['initial_diagnosis']) 

#0    ISM
#1     ISM
#2    WDSM
#3     NaN
#4     ISM
#5     SSM
#6      CM
#7     ASM
#8     ISM 
#Name: initial_diagnosis, dtype: object

missing_values = {nan}
data["initial_diagnosis"] = data["initial_diagnosis"].apply(lambda x: x if x =="ISM" or x in missing_values else "non ISM")
print(data['initial_diagnosis'])

#0    non ISM
#1        ISM
#2    non ISM
#3        NaN
#4        ISM
#5    non ISM
#6    non ISM
#7    non ISM
#8        ISM

不要将 lambda 函数应用于缺失值

Do not apply lambda function to missing values

python

apply

pandas