% FFT_PARTIAL apply fft only on smaller blocks (important for GPU) % % x = fft_partial(x,fft_axis,split_axis, split, inverse = false) % % Inputs: % **x - input array % **fft_axis - axis along which is performed FFT % **split_axis - axis along which is the array split % % *optional* % **split - number of blocks to split the array before FFT to save the memory % **inverse - if true, use ifft intead of fft % % *returns* % ++x fft transformed array %*-----------------------------------------------------------------------* %|                                                                       | %|  Except where otherwise noted, this work is licensed under a          | %|  Creative Commons Attribution-NonCommercial-ShareAlike 4.0            | %|  International (CC BY-NC-SA 4.0) license.                             | %|                                                                       | %|  Copyright (c) 2017 by Paul Scherrer Institute (http://www.psi.ch)    | %|                                                                       | %|      Author: CXS group, PSI  | %*-----------------------------------------------------------------------* % You may use this code with the following provisions: % % If the code is fully or partially redistributed, or rewritten in another % computing language this notice should be included in the redistribution. % % If this code, or subfunctions or parts of it, is used for research in a % publication or if it is fully or partially rewritten for another % computing language the authors and institution should be acknowledged % in written form in the publication: “Data processing was carried out % using the “cSAXS matlab package” developed by the CXS group, % Paul Scherrer Institut, Switzerland.” % Variations on the latter text can be incorporated upon discussion with % the CXS group if needed to more specifically reflect the use of the package % for the published work. % % A publication that focuses on describing features, or parameters, that % are already existing in the code should be first discussed with the % authors. % % This code and subroutines are part of a continuous development, they % are provided “as they are” without guarantees or liability on part % of PSI or the authors. It is the user responsibility to ensure its % proper use and the correctness of the results. function x = fft_partial(x,fft_axis,split_axis, split, inverse) persistent current_gpu import math.* if nargin < 5 inverse = false; % do inverse fft end if nargin < 4 || isempty(split) % auto estimation of the split mem_req = numel(x)*8 * log2(size(x,fft_axis)); if isa(x,'gpuArray') % move directly to complex to include the expected memore requirements x = complex(x); if isempty(current_gpu) || isnan(current_gpu.AvailableMemory) current_gpu = gpuDevice; end split = ceil(mem_req / current_gpu.AvailableMemory); else if mem_req < 50e9 % assume to have at least 50GB ram free ... split = 1; else avail_mem = utils.check_available_memory * 1e6; split = ceil(mem_req /avail_mem); end end end if all(split == 1) if inverse x = ifft(x,[],fft_axis); else x = fft(x,[],fft_axis); end return end % check GPU memory Np = size(x); Nps = Np; Nps(split_axis) = ceil(Nps(split_axis) / split); ind = {':', ':',':'}; for i = 1:split ind{split_axis} = (1+(i-1)*Nps(split_axis)): min(Np(split_axis), i*Nps(split_axis)); x_tmp = x(ind{:}); if ~inverse x_tmp = fft(x_tmp, [], fft_axis); % avoid additonal memory assignment else x_tmp = ifft(x_tmp, [], fft_axis); end x(ind{:}) = x_tmp; end end