Struct fft_state_t¶

#include <fft_core.h>

Public Attributes¶

Type	Name
size_t	n
pocketfft_plan *	plan_f32
pocketfft_plan *	plan_f64
int	sign
double _Complex *	work_trunc

Public Attributes Documentation¶

variable n¶

size_t fft_state_t::n;

Transform length (samples).

variable plan_f32¶

pocketfft_plan* fft_state_t::plan_f32;

CF32 1-D plan.

variable plan_f64¶

pocketfft_plan* fft_state_t::plan_f64;

CF64 1-D plan.

variable sign¶

int fft_state_t::sign;

-1 forward, +1 inverse.

variable work_trunc¶

double _Complex* fft_state_t::work_trunc;

Scratch for a short out. A pocketfft plan is fixed at n and writes all n bins, so it cannot be pointed at a smaller buffer; a truncating call transforms here and copies the prefix. Sized for the widest element (CF64) so one buffer serves both plans. Allocated lazily max_out is n on every sized call, which is everything the Python binding and every in-tree caller does.

The documentation for this class was generated from the following file native/inc/fft/fft_core.h