Module fftw_plan_cache 

FFTW Plan Cache for performance optimization

Caches FFTW plans to avoid expensive plan creation on repeated FFT calls. Plans are stored in global caches keyed by size (1D) or dimensions (2D).

Note: Plans require &mut self for execution, so we use Mutex to ensure exclusive access. This is fine for Python bindings where the GIL serializes calls.

Functions

clear_all_caches

Clear all cached plans (useful for testing or memory management)

execute_c2c_2d_backward

Execute 2D C2C backward FFT with cached plan

execute_c2c_2d_forward

Execute 2D C2C forward FFT with cached plan

execute_c2c_backward

Execute C2C backward (inverse) FFT with cached plan

execute_c2c_forward

Execute C2C forward FFT with cached plan

execute_c2r

Execute C2R (inverse real) FFT with cached plan

execute_c2r_2d

Execute 2D C2R FFT with cached plan

execute_dct2

Execute DCT-II with cached plan

execute_dst2

Execute DST-II with cached plan

execute_idct2

Execute IDCT-II (DCT-III) with cached plan

execute_idst2

Execute IDST-II (DST-III) with cached plan

execute_r2c

Execute R2C FFT with cached plan

execute_r2c_2d

Execute 2D R2C FFT with cached plan