How to convert a 2d numpy array into a 1d numpy array by summing the values and not using for loop?

Question

Is there a numpy function which can combine a 2d numpy array into a 1d numpy array. I want to do it without using a for loop.

Example:

[[1 0 0 0 0], [0 1 0 0 0]] => [1 1 0 0 0]

Answer 1

Just use the ndarray method sum along row axis:

arr2d = np.array([[1, 3, 8, 2, 0], [0, 1, 0, 5, 1]])

arr1d = arr2d.sum(axis=0)

>>> array([1, 4, 8, 7, 1])