2

I have a numpy array, A of size nx1 where each value is a number between 0 and 9.

I would like to create a new array, B of size nx10 such that in B[i] we store a numpy array that contains zeros and a 1 in position A[i].

For example:

A array
[[9]
 [2]
 [4]
 [1]
 [8]]

B array
[[0, 0, 0, 0, 0, 0, 0, 0, 0, 1]
 [0, 0, 1, 0, 0, 0, 0, 0, 0, 0]
 [0, 0, 0, 0, 1, 0, 0, 0, 0, 0]
 [0, 1, 0, 0, 0, 0, 0, 0, 0, 0]
 [0, 0, 0, 0, 0, 0, 0, 0, 1, 0]]

Is there an elegant way of doing this with numpy?

Improve this question
0

3 Answers 3

Reset to default
3

Create a new empty array using numpy.zeros, its size is going to be (arr.size, arr.max()), now fill the items on those positions using multi-dimensional indexing:

>>> arr = np.array([[9], [2], [4], [1], [8]])
>>> arr_ = np.zeros((arr.size, arr.max()))
>>> arr_[np.arange(arr.size), arr[:,0]-1] = 1
>>> arr_
array([[ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],
       [ 0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.],
       [ 1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],
       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.,  0.]])
Improve this answer
Sign up to request clarification or add additional context in comments.

Comments

0

Something like this would do it:

A = [[9],[2],[4],[1],[8]]
B = [[1 if i == j[0] else 0 for i in range(10)] for j in A]

This is a list-based approach; you can simply use np.asarray on B to get the numpy matrix, or create a 10x10 matrix of zeros in numpy and fill 1s in the positions dictated by the A array.

The latter generalises to the case where A's elements might have more than one item.

Improve this answer

Comments

0

You could do something like this, assuming an NP Nx1 array of X:

max_val = max(X)
length_arr = len(X)
new_arr = np.zeros((max_val,length_arr))

This will create the array of the right size that you want.

for i in range(len(X)):
    new_arr[i][X[i]-1]=1

Should then assign the correct values in place?

Works fine in my test case.

Improve this answer

Comments

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.