Question

我正在尝试移动我的一些行并使它们成为列，但保持数据帧的大部分相同。

我的Dataframe看起来像这样：

ID  Thing      Level1 Level2  Time OAttribute IsTrue Score Value
1   bicycle    value  value  9:30  whatever    yes   1     type1
2   non-cycle  value  value  1:30  whatever    no    2     type2
3   bicycle    value  value  2:30  whatever    yes               
4   bicycle    value  value  3:30  whatever    no    4     type3

我想要这样的事情：

ID  Thing Level1 Level2  Time OAttribute      IsTrue  Type1  Type2 Type 3
1   bicycle    value  value  9:30  whatever     yes   1     
2   non-cycle  value  value  1:30  whatever     no            2
3   bicycle    value  value  2:30  whatever      yes               
4   bicycle    value  value  3:30  whatever     no                   4

Answer 1

这样的东西？

In [112]: df
Out[112]: 
   ID      Thing Level1 Level2  Time OAttribute IsTrue  Score  Value
0   1    bicycle  value  value  9:30   whatever    yes      1  type1
1   2  non-cycle  value  value  1:30   whatever     no      2  type2
2   3    bicycle  value  value  2:30   whatever    yes    NaN    NaN
3   4    bicycle  value  value  3:30   whatever     no      4  type3

In [113]: dg = pd.DataFrame(columns=df['Value'].dropna().unique())

In [114]: for i in range(len(df)):
     ...:     key = df.loc[i]['Value']
     ...:     value = df.loc[i]['Score']
     ...:     if key is not pd.np.nan:
     ...:         dg.loc[i, key] = value
     ...:                                                 

In [115]: dg
Out[115]: 
Value type1 type2 type3
0         1   NaN   NaN
1       NaN     2   NaN
3       NaN   NaN     4

In [116]: df.join(dg).drop('Value', 1).fillna('')
Out[116]: 
   ID      Thing Level1 Level2  Time OAttribute IsTrue Score type1 type2 type3
0   1    bicycle  value  value  9:30   whatever    yes     1     1            
1   2  non-cycle  value  value  1:30   whatever     no     2           2      
2   3    bicycle  value  value  2:30   whatever    yes                        
3   4    bicycle  value  value  3:30   whatever     no     4                 4

Answer 2

选项1
使用merge

df_ = df[['ID', 'Value']].dropna().set_index('Value', append=True).ID.unstack()

df.drop('Value', 1).merge(df_, right_index=True, left_index=True, how='left').fillna('')

选项2
使用pd.concat

df_ = df[['ID', 'Value']].dropna().set_index('Value', append=True).ID.unstack()

pd.concat([df, df_], axis=1).drop('Value', 1).fillna('')

将行转换为列，同时保持数据帧的一部分

2 个答案: