Fourier Space Interface

This module contains low level functions for operating in the frequency domain, including:

ParallelKDE.FourierSpace.fourier_statistics! — Function
fourier_statistics!(device, sk, s2k, plan)
fourier_statistics!(device, sk, plan)

Calculate the Fourier transform of kernel first and second moments and its bootstrap samples.

device specifies the device type, which can be :serial, :threaded, or :cuda. sk and s2k are the complex arrays containing the first and second moments of the kernels. plan is the FFTW plan for the Fourier transform.

ParallelKDE.FourierSpace.propagate_statistics! — Function
propagate_statistics!(device, means_t, variances_t, means_0, variances_0, time_propagated, time_initial, grid_array)
propagate_statistics!(device, means_t, means_0, time_propagated, grid_array)

Propagate the first and second moments of the kernels to larger bandwidths.

Arguments

  • Val(device): Specifies the device type, which can be :serial, :threaded, or :cuda.
  • means_t: The complex array where the propagated first moments of the kernels will be stored.
  • means_0: The complex array containing the initial first moments of the kernels.
  • variances_t: The complex array where the propagated second moments of the kernels will be stored.
  • variances_0: The complex array containing the initial second moments of the kernels.
  • time_propagated: A vector of real numbers representing the time points at which the kernels are propagated.
  • time_initial: A vector of real numbers representing the initial time points of the kernels.
  • grid_array: An array of real numbers representing the grid points in the Fourier space.
ParallelKDE.FourierSpace.ifourier_statistics! — Function
ifourier_statistics!(device, sk, s2k, ifft_plan)
ifourier_statistics!(device, sk, ifft_plan)

Calculate the inverse Fourier transform of kernel first and second moments and its bootstrap samples.