python - 根据其他列 pandas 填充缺失值

Question

这个问题已经有答案了:

Pandas conditional creation of a series/dataframe column (13 个回答)

已关闭 4 年前。

我想根据 pandas 中的其他列填充缺失值。这是我的表格:

Gender     Married
Male       Yes
Male       Yes
Female     No
Female     No
Male       NaN
Female     NaN

如果性别是男性，我将填写“已婚”字段的缺失值 ->“已婚”为"is"，否则“已婚”为“否”:

df['Married'].fillna(df[df['Married'].isnull()].apply(lambda x: 'Yes' if (df[df['Married'].isnull()]['Gender'] is 'Male') else 'No', axis=1), inplace=True)

但是失败了，我尝试了很多方法，但没有得到我期望的结果。希望收到大家的来信。

Answer 1

我相信你需要map仅在过滤行中使用 dictionary:

mask = df['Married'].isnull()
df.loc[mask, 'Married'] = df.loc[mask, 'Gender'].map({'Male':'Yes', 'Female':'No'})
print (df)
   Gender Married
0    Male     Yes
1    Male     Yes
2  Female      No
3  Female      No
4    Male     Yes
5  Female      No

另一个解决方案 numpy.where :

mask = df['Married'].isnull()
df.loc[mask, 'Married'] = np.where(df.loc[mask, 'Gender']  == 'Male', 'Yes','No')
print (df)
   Gender Married
0    Male     Yes
1    Male     Yes
2  Female      No
3  Female      No
4    Male     Yes
5  Female      No

另一个解决方案 fillna映射的系列:

df['Married'] = df['Married'].fillna(df['Gender'].map({'Male':'Yes', 'Female':'No'}))