python - 用向量 [n,] 填充逻辑矩阵 [r,n]

Question

我有一个数值向量值( Pandas 数据框df中的一系列)。

idx     values

0          NaN
1            1
2            2
3          NaN
4          NaN
5           33
6           34
7           90
8          NaN
9            5
10         NaN
11          22
12          70
13         NaN
14         672
15          10
16          73
17           9
18         NaN
19          15

然后我构造了一个逻辑矩阵的形式

array([[1, 1, 1, ..., 0, 0, 0],
       [0, 1, 1, ..., 0, 0, 0],
       [0, 0, 1, ..., 0, 0, 0],
       ...,
       [0, 0, 0, ..., 1, 0, 0],
       [0, 0, 0, ..., 1, 1, 0],
       [0, 0, 0, ..., 1, 1, 1]])

使用从 SO 上的一些答案中获取的以下代码，遗憾的是找不到了。

n=len(df)
k=5
r= n-k+1
mat=np.tile([1]*k+[0]*r, r)[:-r].reshape(r,n)

mat 的形状为 (r,n) 而 df['values'] 的形状为 (n,) 。

用 df['values'] 中的值填充 mat 的正确方法是什么？

鉴于前面的例子，我的预期输出是:

array([[NaN, 1, 2, NaN,       ..., 0, 0, 0],
       [  0, 1, 2,NaN,NaN,    ..., 0, 0, 0],
       [  0, 0, 2,NaN,NaN,33, ..., 0, 0, 0],
       ...,
       [0, 0, 0, ...,  672, 10, 73, 9, 0, 0],
       [0, 0, 0, ...,      10,73, 9, NaN, 0],
       [0, 0, 0, ...,        73, 9, NaN, 15]])

关于如何实现这一点有什么建议吗？我尝试了一个点积(希望它会像在 matlab 中一样运行并复制我的向量 r 次但没有成功。

Answer 1

您可以使用 numpy.apply_along_axis和 numpy.where :

#!/usr/bin/env python3

import numpy as np
import pandas as pd

nan = np.nan

df = pd.DataFrame([
         nan, 1, 2, nan, nan, 33, 34, 90, 
         nan, 5, nan, 22, 70, nan, 672, 
         10, 73, 9, nan, 15], 
     columns=['values'])

n = len(df)
k = 5
r = n - k + 1

mat = np.tile([1] * k + [0] * r, r)[:-r].reshape(r, n)

mat = np.apply_along_axis(lambda row: np.where(row, df['values'], row), 1, mat)

print(mat)

输出:

[[ nan   1.   2.  nan  nan   0.   0.   0.   0.   0.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   1.   2.  nan  nan  33.   0.   0.   0.   0.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   2.  nan  nan  33.  34.   0.   0.   0.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.  nan  nan  33.  34.  90.   0.   0.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.  nan  33.  34.  90.  nan   0.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.  33.  34.  90.  nan   5.   0.   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.  34.  90.  nan   5.  nan   0.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.  90.  nan   5.  nan  22.   0.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.  nan   5.  nan  22.  70.   0. 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   5.  nan  22.  70.  nan 0.     0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  nan  22.  70.  nan 672.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  22.  70.  nan 672.  10.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  70.  nan 672.  10.  73.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.  nan 672.  10.  73.   9.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 672.  10.  73.   9.  nan   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 0.    10.  73.   9.  nan  15.]]