Comment créer une matrice de motifs répétitifs. Quand vous voulez faire ce qu'on appelle repmat dans Matlab

Il existe deux méthodes.

Utilisez numpy.matlib.repmat
Utilisez numpy.tile

1. Comment utiliser numpy.matlib.repmat

numpy.matlib.repmat(a, m, n)

In [1]: import numpy as np

In [2]: import numpy.matlib

In [3]: np.matlib.repmat((1,2,3),1,5)
Out[3]: array([[1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3]])

In [4]: A = np.diag((100,200,300))

In [5]: print A
[[100   0   0]
 [  0 200   0]
 [  0   0 300]]

In [6]: np.matlib.repmat(A,2,2)
Out[6]: 
array([[100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300],
       [100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300]])

In [7]: np.matlib.repmat(A,1,2)
Out[7]: 
array([[100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300]])

In [8]: np.matlib.repmat(A,2,1)
Out[8]: 
array([[100,   0,   0],
       [  0, 200,   0],
       [  0,   0, 300],
       [100,   0,   0],
       [  0, 200,   0],
       [  0,   0, 300]])

2. Comment utiliser numpy.tile

numpy.tile(a, reps)

In [9]: np.tile((1,2,3),(1,5))
Out[9]: array([[1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3]])

In [10]: np.tile(A,(2,2))
Out[10]: 
array([[100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300],
       [100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300]])

In [11]: np.tile(A,(1,2))
Out[11]: 
array([[100,   0,   0, 100,   0,   0],
       [  0, 200,   0,   0, 200,   0],
       [  0,   0, 300,   0,   0, 300]])

In [12]: np.tile(A,(2,1))
Out[12]: 
array([[100,   0,   0],
       [  0, 200,   0],
       [  0,   0, 300],
       [100,   0,   0],
       [  0, 200,   0],
       [  0,   0, 300]])

Si vous utilisez repmat, par exemple, vous pouvez l'utiliser lorsque vous souhaitez ne laisser que les positions paires du tableau bidimensionnel et définir le reste sur 0.

In [27]: X = np.arange(100).reshape(10,10)

In [28]: Y = np.matlib.repmat(np.array([[1,0],[0,0]]),5,5)

In [29]: print X
[[ 0  1  2  3  4  5  6  7  8  9]
 [10 11 12 13 14 15 16 17 18 19]
 [20 21 22 23 24 25 26 27 28 29]
 [30 31 32 33 34 35 36 37 38 39]
 [40 41 42 43 44 45 46 47 48 49]
 [50 51 52 53 54 55 56 57 58 59]
 [60 61 62 63 64 65 66 67 68 69]
 [70 71 72 73 74 75 76 77 78 79]
 [80 81 82 83 84 85 86 87 88 89]
 [90 91 92 93 94 95 96 97 98 99]]

In [30]: print Y
[[1 0 1 0 1 0 1 0 1 0]
 [0 0 0 0 0 0 0 0 0 0]
 [1 0 1 0 1 0 1 0 1 0]
 [0 0 0 0 0 0 0 0 0 0]
 [1 0 1 0 1 0 1 0 1 0]
 [0 0 0 0 0 0 0 0 0 0]
 [1 0 1 0 1 0 1 0 1 0]
 [0 0 0 0 0 0 0 0 0 0]
 [1 0 1 0 1 0 1 0 1 0]
 [0 0 0 0 0 0 0 0 0 0]]

In [31]: Z = X * Y

In [32]: print Z
[[ 0  0  2  0  4  0  6  0  8  0]
 [ 0  0  0  0  0  0  0  0  0  0]
 [20  0 22  0 24  0 26  0 28  0]
 [ 0  0  0  0  0  0  0  0  0  0]
 [40  0 42  0 44  0 46  0 48  0]
 [ 0  0  0  0  0  0  0  0  0  0]
 [60  0 62  0 64  0 66  0 68  0]
 [ 0  0  0  0  0  0  0  0  0  0]
 [80  0 82  0 84  0 86  0 88  0]
 [ 0  0  0  0  0  0  0  0  0  0]]

Dans le cas d'une image couleur, il s'agit de données tridimensionnelles, c'est donc comme suit.

`ipython`


In [1]: from skimage import data

In [2]: from skimage import io

In [3]: import numpy as np

In [4]: sz = 48

In [5]: image = data.lena()

In [6]: patch1 = image[250:250+sz,250:250+sz,:]

In [7]: patch1.shape
Out[7]: (48, 48, 3)

In [8]: mask = np.tile(np.array([[[1],[0]],[[0],[0]]],dtype=np.uint8),(sz/2,sz/2,3))

In [9]: mask.shape
Out[9]: (48, 48, 3)

In [10]: print mask
[[[1 1 1]
  [0 0 0]
  [1 1 1]
  ..., 
  [0 0 0]
  [1 1 1]
  [0 0 0]]


 [[0 0 0]
  [0 0 0]
  [0 0 0]
  ..., 
  [0 0 0]
  [0 0 0]
  [0 0 0]]]

In [11]: patch2 = patch1 * mask

In [12]: patch12 = np.hstack((patch1,patch2))

[Python] Comment créer une matrice de motifs répétitifs (repmat / tile)

1. Comment utiliser numpy.matlib.repmat

2. Comment utiliser numpy.tile

ipython

`ipython`