Replies: 2 comments
-
I just edited the question to have more consistent notation. |
Beta Was this translation helpful? Give feedback.
0 replies
-
Hi -
The cost is usual FFT plus linear in number of NU points - this is in the refs.
…Sent from my iPod
On Sep 20, 2025, at 7:04 AM, Kaya Unalmis ***@***.***> wrote:
Closed #731 as resolved.
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Let$k_x$ , $k_y$ be the number of complex Fourier modes in each direction. The asymptotic complexity of a 2D FFT from $k_x \times k_y$ frequency space grid onto denser $p_x \times p_y$ points grid with $p = p_x \times p_y \gg \min(k_x, k_y)$ reduces to
$$\mathcal{O}(p \log p)$$
What is the complexity of FINUFFT's implementation of the 2D type 2 NUFFT onto p non-uniform points?
$$\mathcal{O}[p \log(p) + k_x k_y]$$
A naive extrapolation of the complexity listed for the 1D type 1 NUFFT would imply
Is that the scaling or is it like
$$\mathcal{O}[p \log(p) \log(k_x k_y) + k_x k_y]$$
I could not find this information in the published references.
Beta Was this translation helpful? Give feedback.
All reactions