[FIXED] Getting the values of only one channel of a numpy array

Issue

I would very much be grateful if you don't close this question without even giving a hint of how to solve this problem,please.

I have the following

import numpy as np

layer1 = np.zeros((5,3,4),dtype=np.uint8)
layer1[0,0,0]=20
layer1[1,1,0]=20
layer1[2,2,0]=20
layer1[3,1,0]=20
layer1[4,0,0]=20

layer1[0,0,1] =50
layer1[1,0,1]=50
layer1[2,0,1]=50


print(layer1)
print("---------------")

which gives me

    [[[20 50  0  0]
  [ 0  0  0  0]
  [ 0  0  0  0]]

 [[ 0 50  0  0]
  [20  0  0  0]
  [ 0  0  0  0]]

 [[ 0 50  0  0]
  [ 0  0  0  0]
  [20  0  0  0]]

 [[ 0  0  0  0]
  [20  0  0  0]
  [ 0  0  0  0]]

 [[20  0  0  0]
  [ 0  0  0  0]
  [ 0  0  0  0]]]

How can I reduce a get the values only of one channel ?

For example for channel=0

I want to get

[[20  0  0]
 [ 0 20  0]
 [ 0  0 20]
 [ 0 20  0]
 [20  0  0]]

where channel can be 0,1,2 or 3

EDIT: Just in case, the layer1[0,0,0]=20 is just a convenient way to fill up the matrix. My question is how to tranform layer1 once filled to the matrix of (5,3)

EDIT: if the "channel" is 1 then I would get

[[50  0  0]
 [ 50 0  0]
 [ 50  0 0]
 [ 0 0  0]
 [0  0  0]]

Solution

numpy array indexing is well documented. Don't skip it!

In [1]: layer1 = np.zeros((5,3,4),dtype=np.uint8)
   ...: layer1[0,0,0]=20
   ...: layer1[1,1,0]=20
   ...: layer1[2,2,0]=20
   ...: layer1[3,1,0]=20
   ...: layer1[4,0,0]=20
   ...: 
   ...: layer1[0,0,1] =50
   ...: layer1[1,0,1]=50
   ...: layer1[2,0,1]=50
In [2]: layer1.shape
Out[2]: (5, 3, 4)
In [3]: layer1[:,:,0]
Out[3]: 
array([[20,  0,  0],
       [ 0, 20,  0],
       [ 0,  0, 20],
       [ 0, 20,  0],
       [20,  0,  0]], dtype=uint8)
In [4]: layer1[:,:,2]
Out[4]: 
array([[0, 0, 0],
       [0, 0, 0],
       [0, 0, 0],
       [0, 0, 0],
       [0, 0, 0]], dtype=uint8)

Answered By - hpaulj