How to replace each element in a numpy array with multiple values?

I have a 2D numpy array with size n * m, and I need to replace each value with 2 valus and result in an array with size n * 2m.

Replace pattern:1 with [1,0], 2 with [0,1] and 0 with [0,0]

Input: [[1,0,2],[2,2,1]]

Desired Output: [[1,0,0,0,0,1],[0,1,0,1,1,0]]

It can be easily done with a for loop, however I’m trying to find a ‘numpy’ way of doing it, to achieve maximum efficiency.

res = np.zeros((arr.shape[0],arr.shape[1]*2))
for index, row in enumerate(arr):
    temp = np.array([])
    for e in row:
        if e == 1:
            temp = np.append(temp,[1,0])
        elif e == 2:
            temp = np.append(temp,[0,1])
        else:
            temp = np.append(temp,[0,0])
    res[index] = temp

>Solution :

As you have values 0, 1, 2, I would use a reference array to map the values by indexing:

a = np.array([[1,0,2],[2,2,1]])
mapper = np.array([[0, 0],  # position 0
                   [1, 0],  # position 1
                   [0, 1]]) # position 2

out = mapper[a].reshape((len(a), -1))

Output:

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