如何制作一个数组，其中每个元素都是随机图像？

Question

I want to make an array in python that has about 40000 elements and each element is a random image with size 28x28. I use this code but it produces the following error. I am a beginner in python.

    wtrain=np.zeros((40000,28,28,1))
    for i in range(40000):
        w_main = np.random.randint(2,size=(1,4,4,1))
        w_main=w_main.astype(np.float32)
        w_expand=np.zeros((1,28,28,1),dtype='float32')
        w_expand[:,0:4,0:4]=w_main
        w_expand.reshape(1,28,28,1)
        wtrain[i,:,:,:]=w_expand

the error

All input arrays (x) should have the same number of samples. Got array shapes: [(49999, 28, 28, 1), (1, 28, 28, 1)]

what is the problem? how can I add these random images to wtrain? Thanks

I change my code to this:

wtrain=[]
for i in range(2):
    w_main = np.random.randint(2,size=(1,4,4,1))
    w_main=w_main.astype(np.float32)
    w_expand=np.zeros((1,28,28,1),dtype='float32')
    w_expand[:,0:4,0:4]=w_main
    w_expand.reshape(1,28,28,1)
    wtrain.append(w_expand)

Answer 1

Make a zeros array of the required shape - I used (3,7,7) instead of your (40000,28,28).

a = np.zeros((3,7,7))

Make an array of random numbers with a shape of your spec - the first dimension size is the same as the zeros array

b = np.random.randint(0,255, size=(3,4,4))

Assign the random values to the zeros array using indexing on the left-hand-side of the assignment to put those values where you want

a[:,:4,:4] = b

Result

In [20]: a[0,...]
Out[20]: 
array([[ 241.,  228.,  176.,  194.,    0.,    0.,    0.],
       [ 185.,  240.,  219.,  175.,    0.,    0.,    0.],
       [ 206.,   82.,   32.,  137.,    0.,    0.,    0.],
       [  58.,  181.,  242.,  168.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.]])

In [21]: a[1,...]
Out[21]: 
array([[  25.,   19.,  251.,   89.,    0.,    0.,    0.],
       [ 204.,   25.,   72.,  176.,    0.,    0.,    0.],
       [ 189.,   37.,   33.,   49.,    0.,    0.,    0.],
       [  72.,  168.,   68.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.]])

In [22]: a[2,...]
Out[22]: 
array([[  74.,  228.,  186.,  133.,    0.,    0.,    0.],
       [ 147.,   83.,  194.,  205.,    0.,    0.,    0.],
       [  34.,  185.,   21.,    6.,    0.,    0.,    0.],
       [  14.,  245.,   46.,  154.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.],
       [   0.,    0.,    0.,    0.,    0.,    0.,    0.]])

Add another dimension by reshaping

In [26]: x,y,z = a.shape

In [27]: c = a.reshape((x,y,z,1))

In [28]: c.shape
Out[28]: (3, 7, 7, 1)