Clear Filters
Clear Filters

Load and Save using PARFOR Loops

19 views (last 30 days)
Anas Khan
Anas Khan on 25 Oct 2022
Commented: Anas Khan on 25 Oct 2022
I am using PARFOR loops on a cluster like so:
files = dir('*.mat'); % Get list of data files to load
subjectIDs = cell(1,length(files));
% Extract subject name from each filename (SUBJECTNAME.mat format for each file)
for file = 1:length(files)
subidsplit = strsplit(files(file).name,'.');
subjectIDs{file} = subidsplit{1};
end
clear file subidsplit
frex = 2.^([10:76]/10);
parpool(length(files),'AttachedFiles',"wilson_sf.m") % Initialize parpool,
% one worker for each iteration, attach helper function
parfor ii = length(files)
CheahaGGC(1000, frex, 0, 0, 6,files,ii,subjectIDs);
end
Inside the function CheahaGGC, I am loading one file, running analysis on it, and saving it. My script completed without errors in 7 hours, but only one of the iteration's results were saved (the last one, corresponding to ii = length(files)). Is there a problem with how I implemented the loop? Each worker gets its own workspace so calling LOAD inside my function should give each worker its own data to analyze right? Same thing with SAVE. Shown below is the load and save inside of the function
function CheahaGGC(fs, freq, pad0, flgCond, omega0,files,ii,subjectIDs)
load(files(ii).name) % Load one file for this iteration
%%%%%%%%%%%%%%%%%%%%%%%
%%%% Analysis Code %%%%
%%%%%%%%%%%%%%%%%%%%%%%
save(strcat(subjectIDs{ii},'GGC','.mat'),'causalityGO','causalityNG') % Save results to file with subject name + "GGC" added

Accepted Answer

Bruno Luong
Bruno Luong on 25 Oct 2022
Edited: Bruno Luong on 25 Oct 2022
Normal
parfor ii = length(files)
should be
parfor ii = 1:length(files)
  1 Comment
Anas Khan
Anas Khan on 25 Oct 2022
Wow... thanks. Cannot believe the rabit holes I went down when it was that simple.

Sign in to comment.

More Answers (0)

Categories

Find more on Parallel for-Loops (parfor) in Help Center and File Exchange

Products


Release

R2021b

Community Treasure Hunt

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

Start Hunting!