python - 根据索引选择固定数据帧范围并附加到新数据帧

Question

我有一个数据框:

resultsDf

返回以下内容:

       0
0     100
1   -2800
2   -2800
3   -2800
0   -2800
1   -2800
2   -2900
3   -3000
0   -3000
1   -3000
2   -3000
3   -3000
0   -3000
1   -3000
2   -3000
3   -3000
.  
.  
.   
0  -3100
1  25500

我想根据索引提取数据帧的固定子集，即 0,1,2,3

然后我想将每个帧添加到列格式的新数据框中。所以最终数据框应该是这样的:

  C1   C2   C3.....Cn
0
1
2
3

Answer 1

你可以通过 values 将 df 转换为 numpy 数组和 reshape .然后您可以使用 range 通过列表理解设置新的列名称:

print resultsDf

      0
0   100
1 -2800
2 -2800
3 -2800
0 -2800
1 -2800
2 -2900
3 -3000
0 -3000
1 -3000
2 -3000
3 -3000
0 -3000
1 -3000
2 -3000
3 -3000

df = pd.DataFrame((resultsDf.values).reshape((4, (resultsDf.values).shape[0]/4)))
df.columns = ['C' + str(i) for i in range(1, len(df.columns) + 1) ]

print df

     C1    C2    C3    C4
0   100 -2800 -2800 -2800
1 -2800 -2800 -2900 -3000
2 -3000 -3000 -3000 -3000
3 -3000 -3000 -3000 -3000

如果您缺少最后一行(没有像其他行那样重复索引):

print resultsDf

        0
0     100
1   -2800
2   -2800
3   -2800
4   -2800
5   -2800
6   -2900
7   -3000
8   -3000
9   -3000
10  -3000
11  -3000
12  -3000
13  -3000
14  -3000
15  -3000
0     100
1   -2800
2   -2800
3   -2800
4   -2800
5   -2800
6   -2900
7   -3000
8   -3000
9   -3000
10  -3000
11  -3000
12  -3000
13  -3000
14  -3000
15  -3000
0   -3100
1   25500

#use all df without last two rows - resultsDf[:-2]
df = pd.DataFrame((resultsDf[:-2].values).reshape(16, resultsDf[:-2].values.shape[0]/16))
#append last two rows to new df
df = pd.concat([df, resultsDf[-2:]], axis=1)
df.columns = ['C' + str(i) for i in range(1, len(df.columns) + 1) ]

print df

      C1    C2     C3
0    100 -2800  -3100
1  -2800 -2800  25500
2  -2800 -2800    NaN
3  -2900 -3000    NaN
4  -3000 -3000    NaN
5  -3000 -3000    NaN
6  -3000 -3000    NaN
7  -3000 -3000    NaN
8    100 -2800    NaN
9  -2800 -2800    NaN
10 -2800 -2800    NaN
11 -2900 -3000    NaN
12 -3000 -3000    NaN
13 -3000 -3000    NaN
14 -3000 -3000    NaN
15 -3000 -3000    NaN