Struct BluesteinsAlgorithm 

pub struct BluesteinsAlgorithm<T> { /* private fields */ }

Implementation of Bluestein’s Algorithm

This algorithm computes an arbitrary-sized FFT in O(nlogn) time. It does this by converting this size-N FFT into a size-M FFT where M >= 2N - 1.

The choice of M is very important for the performance of Bluestein’s Algorithm. The most obvious choice is the next-largest power of two – but if there’s a smaller/faster FFT size that satisfies the >= 2N - 1 requirement, that will significantly improve this algorithm’s overall performance.

// Computes a forward FFT of size 1201, using Bluestein's Algorithm
use rustfft::algorithm::BluesteinsAlgorithm;
use rustfft::{Fft, FftPlanner};
use rustfft::num_complex::Complex;

let mut buffer = vec![Complex{ re: 0.0f32, im: 0.0f32 }; 1201];

// We need to find an inner FFT whose size is greater than 1201*2 - 1.
// The size 2401 (7^4) satisfies this requirement, while also being relatively fast.
let mut planner = FftPlanner::new();
let inner_fft = planner.plan_fft_forward(2401);

let fft = BluesteinsAlgorithm::new(1201, inner_fft);
fft.process(&mut buffer);

Bluesteins’s Algorithm is relatively expensive compared to other FFT algorithms. Benchmarking shows that it is up to an order of magnitude slower than similar composite sizes. In the example size above of 1201, benchmarking shows that it takes 5x more time to compute than computing a FFT of size 1200 via a step of MixedRadix.

Implementations

impl<T> BluesteinsAlgorithm<T>

where T: FftNum,

pub fn new(len: usize, inner_fft: Arc<dyn Fft<T>>) -> BluesteinsAlgorithm<T>

Creates a FFT instance which will process inputs/outputs of size len. inner_fft.len() must be >= len * 2 - 1

Note that this constructor is quite expensive to run; This algorithm must compute a FFT using inner_fft within the constructor. This further underlines the fact that Bluesteins Algorithm is more expensive to run than other FFT algorithms

Panics

Panics if inner_fft.len() < len * 2 - 1.

Trait Implementations

impl<T> Direction for BluesteinsAlgorithm<T>

where T: FftNum,

fn fft_direction(&self) -> FftDirection

Returns FftDirection::Forward if this instance computes forward FFTs, or FftDirection::Inverse for inverse FFTs

impl<T> Fft<T> for BluesteinsAlgorithm<T>

where T: FftNum,

fn process_immutable_with_scratch( &self, input: &[Complex<T>], output: &mut [Complex<T>], scratch: &mut [Complex<T>], )

Divides input and output into chunks of self.len(), and computes a FFT on each chunk while keeping input untouched.

fn process_outofplace_with_scratch( &self, input: &mut [Complex<T>], output: &mut [Complex<T>], scratch: &mut [Complex<T>], )

Divides input and output into chunks of size self.len(), and computes a FFT on each chunk.

fn process_with_scratch( &self, buffer: &mut [Complex<T>], scratch: &mut [Complex<T>], )

Divides buffer into chunks of size self.len(), and computes a FFT on each chunk.

fn get_inplace_scratch_len(&self) -> usize

Returns the size of the scratch buffer required by process_with_scratch

fn get_outofplace_scratch_len(&self) -> usize

Returns the size of the scratch buffer required by process_outofplace_with_scratch

fn get_immutable_scratch_len(&self) -> usize

Returns the size of the scratch buffer required by process_immutable_with_scratch

fn process(&self, buffer: &mut [Complex<T>])

Computes a FFT in-place.

impl<T> Length for BluesteinsAlgorithm<T>

where T: FftNum,

fn len(&self) -> usize

The FFT size that this algorithm can process