필터 지우기
필터 지우기

How to speed up convolution with a million data points

조회 수: 11 (최근 30일)
Runzi Hao
Runzi Hao 2024년 5월 3일
편집: Runzi Hao 2024년 5월 6일
I am currently doing convolution using nested for loops, for 10^6 data points in each for loop. Are there ways to speed up the following code? Thanks in advance!
% nIters = 40;
% n = 1e6;
% mzL = rand(nIters, n);
% gg = rand(1, n);
mzR_temp = zeros(nIters, 1);
for c = 1:n
mzR_temp(:) = 0;
for d = 1:c
mzR_temp(:) = mzR_temp(:) + gg(c-d+1) * mzL(:,d);
end
mzR_II(:,c) = mzR_temp;
end

채택된 답변

Matt J
Matt J 2024년 5월 3일
편집: Matt J 2024년 5월 3일
Use conv,
mzR_II=conv(gg,mzL,'same');
or FFTs,
mzR_II=ifft( fft(gg,2*n) .* fft(mzL,2*n) , 'symmetric');
mzR_II=mzR_II(1:n);
  댓글 수: 4
Matt J
Matt J 2024년 5월 3일
편집: Matt J 2024년 5월 3일
There is absolutely no way the computation should take more than 1 second on any computer made within the last 10 years.
n=1e6;
mzL = rand(1,n);
gg = rand(1,n-1);
tic
mzR_II = [0 fftfilt(mzL,gg)];
toc
Elapsed time is 0.174552 seconds.
tic;
mzR_II=ifft( fft(gg,2*n) .* fft(mzL,2*n) , 'symmetric');
mzR_II=mzR_II(1:n);
toc
Elapsed time is 0.145911 seconds.
Runzi Hao
Runzi Hao 2024년 5월 6일
Awesome! Thanks for the suggestion of using fftfilt.
It turns out that, on my computer, the convolution ran 43 s for a million data points; and with 40 iterations, the total time was 2600+ s. However, the fftfilt function ran only 8 s for the 40 million data points!

댓글을 달려면 로그인하십시오.

추가 답변 (1개)

Image Analyst
Image Analyst 2024년 5월 3일
Use the built-in convolutions functions: conv and conv2. They are highly optimized for speed.

카테고리

Help CenterFile Exchange에서 Fourier Analysis and Filtering에 대해 자세히 알아보기

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by