Out of memory error

I'm trying to run the code below, it works well and does its job. The problem is that it is not capable of running all iterations continuously, the out of memory error always appears and I end up having to restart it or start it running from where it left off. I've already made some changes, but the error still persists. Does anyone have any suggestions or ideas on how I can get around it?

The algorithm saves one image from each cycle, with each iteration almost 1800 images are generated, I don't know if this is what is causing the error to occur. But using the close(figh) command it closes the image that is used to save...

clear all; close all; clc; warning off;

baseDir = 'C:\Users\Lucas\Documents\Mestrado\wsst';
caso = 2; % Selecionando o caso 2 para simulação
N = 32;
ciclo = 128;
tipos_de_falta = ["A", "B", "C", "AB", "BC", "CA", "ABC"];
hs = [0, 25, 50, 75, 100];
runs = 1:9;

for idx_tipo = 1:length(tipos_de_falta)
    for idx_h = 1:length(hs)
        for idx_run = 1:length(runs)
            clearvars -except baseDir caso N ciclo hs tipos_de_falta runs idx_tipo idx_h idx_run

            tipo_falta = tipos_de_falta(idx_tipo);
            h = hs(idx_h);
            run = runs(idx_run);

            if caso == 2
                [Ip_faseA, Ip_faseB, Ip_faseC, Is_faseA, Is_faseB, Is_faseC, If_faseA, If_faseB, If_faseC, t, instanteChaveamento, prefixo_arquivo, tipo_falta_descricao] = PSCAD_2(h, N, tipo_falta, run);
            else
                disp('Opção inválida. Por favor, escolha 1, 2 ou 3.');
                break;
            end

            pastaCaso = ['Falta_', prefixo_arquivo, '_em_', num2str(h),'_da_LT'];
            casoPath = fullfile(baseDir, pastaCaso);
            if ~exist(casoPath, 'dir')
                mkdir(casoPath);
            end
            
            chaveamentoFolderName = sprintf('Chaveamento_%0.6f_s', instanteChaveamento);
            chaveamentoFolderName = regexprep(chaveamentoFolderName, '[\s\.]', '');
            chaveamentoPath = fullfile(casoPath, chaveamentoFolderName);
            if ~exist(chaveamentoPath, 'dir')
                mkdir(chaveamentoPath);
            end
            
            subFolders = {'pre_falta', 'chaveamento', 'falta'};
            phases = {'Fase A', 'Fase B', 'Fase C'};

            for i = 1:length(subFolders)
                subFolderPath = fullfile(chaveamentoPath, subFolders{i});
                if ~exist(subFolderPath, 'dir')
                    mkdir(subFolderPath);
                end
                
                for j = 1:length(phases)
                    phaseFolderPath = fullfile(subFolderPath, phases{j});
                    if ~exist(phaseFolderPath, 'dir')
                        mkdir(phaseFolderPath);
                    end
                end
            end

            Ida = If_faseA;
            Idb = If_faseB;
            Idc = If_faseC;

            Ida = normalize(Ida);
            Idb = normalize(Idb);
            Idc = normalize(Idc);

            total_a = length(t);
            total_d = t(end) - t(1);
            fs = total_a / total_d;

            for i = 1:(total_a - ciclo + 1)
                i_limite = min(i + ciclo, total_a);
                tmin = t(i);
                tmax = t(i_limite);

                if tmin >= 1
                    break;
                end

                if tmax < instanteChaveamento
                    pasta_destino = 'pre_falta';
                elseif tmin >= instanteChaveamento
                    pasta_destino = 'falta';
                else
                    pasta_destino = 'chaveamento';
                end

                for fase = ['a', 'b', 'c']
                    [sst, f] = wsst(eval(['Id' fase '(i:i_limite)']), fs, 'ExtendSignal', true);
                    figh = figure('Visible', 'off');
                    pcolor(t(i:i_limite), f, abs(sst));
                    shading interp;
                    set(figh, 'Position', get(0, 'Screensize'));
                    set(gca, 'Visible', 'off');
                    colorbar('off');
                    
                    caminho_do_arquivo = sprintf('%s\\%s\\Fase %s\\TF_Ridges_Fase_%s_%s_em_%d_LT_%d.jpg', ...
                        chaveamentoPath, pasta_destino, upper(fase), upper(fase), prefixo_arquivo, h, i);
                    exportgraphics(gca, caminho_do_arquivo);
                    close(figh); % Fechar a figura atual para economizar memória
                end
            end
        end
    end
end

[Walter Roberson]

I do not see any obvious reason.

I am a bit concerned about the use of eval()

evil

[Walter Roberson]

It would be more robust to be explicit which axes you are working with

                    figh = figure('Visible', 'off');
                    ax = axes('parent', figh);
                    pcolor(ax, t(i:i_limite), f, abs(sst));
                    shading(ax, 'interp');
                    set(figh, 'Position', get(0, 'Screensize'));
                    set(ax, 'Visible', 'off');
                    colorbar(ax, 'off');
                    
                    caminho_do_arquivo = sprintf('%s\\%s\\Fase %s\\TF_Ridges_Fase_%s_%s_em_%d_LT_%d.jpg', ...
                        chaveamentoPath, pasta_destino, upper(fase), upper(fase), prefixo_arquivo, h, i);
                    exportgraphics(ax, caminho_do_arquivo);
                    close(figh); % Fechar a figura atual para economizar memória

Answer 1

[Ashutosh Thakur]

Hello Luccas,

The approach of splitting the codebase into the calculation part and the graphics part can be done, and then both sections can be profiled for benchmarking. You can use the memory function in MATLAB to identify memory usage and calculate the memory usage of different operations, thereby identifying the operations that are causing memory issues. You can refer to the following documentation links:

• https://www.mathworks.com/help/matlab/performance-and-memory.html.
• https://www.mathworks.com/help/matlab/ref/memory.html.

To resolve Out of Memory issues, you can follow the following link on the strategies and methods described in the following link:

https://www.mathworks.com/matlabcentral/answers/95781-why-do-i-get-out-of-memory-errors-in-matlab?s_tid=answers_rc1-3_p3_Topic

I hope this helps you!

[Luccas S.]

The error that appears is this:

Error using exportgraphics

Out of memory

exportgraphics(gca, caminho_do_arquivo)