add a new dimension and repeat values for a numpy array

Question

I have a very basic question. I think this can be done elegantly in a single line with numpy, but I don't know how. Let's say I have a 2D numpy array with shape (Nx,Ny), and I want to add a third dimension with shape Nz. For every fixed index_x,index_y I want to repeat the value of the array for all indices across the new z axis.

I can easily do this with basic instructions like

# array_1 has shape (Nx,Ny)
array_2 = np.zeros((Nx,Ny,Nz))

for idx_x in range(Nx):
    for idx_y in range(Ny):
        array_2[idx_x,idx_y,:] = array_1[idx_x,idx_y]

but I was wondering what is the exact numpy function to do this in a single line

reshape, expend_dims or the [:,:,None] idiom can be used to add a dimension. np.repeat (or np.tile) can repeat the values. — hpaulj
– hpaulj, Commented Apr 29, 2023 at 16:04

Demis the Anacronist · Accepted Answer · 2023-04-29 19:40:37Z

0

the logic depend, every plane is the same or is linear?

Must be a cube or something else?

comes from a rectangular shape or squared?

Anyway, use Array.reshape

This is a little example

Array2 = Array1.reshape(len(Array1), len(Array1[0]), <someValue>)

Another source

edited Apr 29, 2023 at 19:40

answered Apr 29, 2023 at 16:49

Demis the Anacronist

1411 silver badge9 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Nick ODell Over a year ago

This isn't syntactically correct - mismatch in number of ( and ).

Demis the Anacronist Over a year ago

hey thank you this was a little misslook

David · Accepted Answer · 2023-04-30 01:33:44Z

You have two options:

As mentioned in the comments, you can add another dimension to the array and repeat along that:

np.repeat(array_1[:, :, None], Nz, axis=2)

A better option is to use np.tile. You don't need to manually add a dimension and it can easily be extended to other cases when you want to repeat along one of the existing dimensions:

np.tile(array_1, (1, 1, Nz))

Basically, you specify how many times you want to repeat the array along each axis.

Collectives™ on Stack Overflow

add a new dimension and repeat values for a numpy array

2 Answers 2

2 Comments

Comments

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Related