python数据帧to_dict按索引，排除NaN

Question

采样数据帧

     v1 v2  v3  v4  v5  v6
index                                                                   
0   -2  -2  -2  NaN -2  -2
1   -2  -2  -2  NaN -2  -2
2   -2  -2  -2  NaN -2  -2
3   -2  -2  -2  -2  -2  -2
4   -2  -2  -2  NaN -2  -2

我正在尝试将数据框更改为字典 我想忽略 json 文档中的“nan”字段

正在尝试什么：

import pandas as pd
pd.DataFrame((samp.to_dict('index')).items())

期望的输出：

index    values
0   {'v1':'-2', 'v2':'-2','v3':'-2','v5':'-2','v6':'-2'}
1   {'v1':'-2', 'v2':'-2','v3':'-2','v5':'-2','v6':'-2'}
2   {'v1':'-2', 'v2':'-2','v3':'-2','v5':'-2','v6':'-2'}
3   {'v1':'-2', 'v2':'-2','v3':'-2','v4':'-2','v5':'-2','v6':'-2'}
4   {'v1':'-2', 'v2':'-2','v3':'-2','v5':'-2','v6':'-2'}

Answer 1

用apply做吧

df.apply(lambda x : x.dropna().to_dict(),axis=1)
Out[362]: 
index
0    {'v3': -2.0, 'v5': -2.0, 'v6': -2.0, 'v1': -2....
1    {'v3': -2.0, 'v5': -2.0, 'v6': -2.0, 'v1': -2....
2    {'v3': -2.0, 'v5': -2.0, 'v6': -2.0, 'v1': -2....
3    {'v4': -2.0, 'v3': -2.0, 'v5': -2.0, 'v2': -2....
4    {'v3': -2.0, 'v5': -2.0, 'v6': -2.0, 'v1': -2....
dtype: object