How to collate mutiple mat files based on their name?

Hi all,

I have multiple folders with mutiple mat files which are structures with a table inside. Please see screenshot below of an exmaple folder and example mat file attached.

I want to:

Collate files ending with the same number (as in green below, these numbers may differ depending on a folder) into one mat file with a name ending with this speciif number (for example: Dot_DatSynced_ID001_001). This mat file should be a structure with tables named LShank, LThigh, Pelvis, RShank, RThigh.

Could you please help?

Answer 1

[ag]

Hi Tomaszzz,

To collate files ending with the same number into one mat file, you can follow the below steps:

• Use MATLAB's dir function to list all .mat files in the folder.
• Extract unique numbers from filenames to know which files to group together.
• Load and Combine Data: For each unique number, load the corresponding files and combine them into a structure.
• Save Combined Data: Save the combined structure into a new ".mat" file.

The below pseudocode demonstrates how to achieve this:

% Get list of files
files = dir('*.mat');

% Extract unique endings
endings = [];
for i = 1:length(files)
    % Extract number ending from filename
    nameParts = split(files(i).name, '_');
    ending = nameParts{end-1};
    endings = [endings; ending];
end
uniqueEndings = unique(endings);

% Process each unique ending
for i = 1:length(uniqueEndings)
    ending = uniqueEndings{i};
    
    % Initialize structure to hold tables
    dataStruct = struct();
    
    % Load files with the same ending
    for j = 1:length(files)
        if contains(files(j).name, ending)
            % Load the mat file
            loadedData = load(files(j).name);
            
            % Determine which table it is (LShank, LThigh, etc.)
            if contains(files(j).name, 'LShank')
                dataStruct.LShank = loadedData.table;
            else ...
            end
        end
    end
    
    % Step 5: Save the combined structure
    saveFileName = ['Dot_DataSynced_ID001_' ending '.mat'];
    save(saveFileName, '-struct', 'dataStruct');
end

For more details, please refer to the following MathWorks documentations:

• dir - https://www.mathworks.com/help/matlab/ref/dir.html
• split - https://www.mathworks.com/help/matlab/ref/string.split.html
• save - https://www.mathworks.com/help/matlab/ref/save.html

Hope this helps!

Answer 2

[Voss]

% first, make some copies of that single mat file, 
% with different names, for demonstration:
F = dir('*.mat');
copyfile(F(1).name,'Dot_DataSynced_ID001_001_RShank_mat.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_001_Pelvis_mat.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_007_LShank_mat.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_007_RShank_math.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_007_RThigh_math.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_011_Pelvis_booey.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_011_baba_booey.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_011_LThigh_krokus.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_015_LLCOOLJ_quivers.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_015_RCRUMB_norris.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_015_richard_christy.mat')
copyfile(F(1).name,'Dot_DataSynced_ID001_015_sal_governale.mat')


% now perform the operations specified in the question:
out_folder = 'Collated';
F = dir('*.mat');
C = regexp({F.name},'(.*)_([^_]*)_([^_]*)_.*','tokens','once');
C = vertcat(C{:});
[G,GID] = findgroups(C(:,2));
if ~isfolder(out_folder)
    mkdir(out_folder)
end
for ii = 1:numel(GID)
    out = struct();
    idx = find(G == ii);
    for jj = 1:numel(idx)
        S = load(fullfile(F(idx(jj)).folder,F(idx(jj)).name));
        tmp = fieldnames(S);
        if isempty(tmp)
            continue
        end
        out.(C{idx(jj),3}) = S.(tmp{1});
    end
    save(fullfile(out_folder,sprintf('%s_%s.mat',C{idx(1),1},GID{ii})),'-struct','out')
end


% now check the resulting mat files
F = dir(fullfile(out_folder,'*.mat'));
for ii = 1:numel(F)
    disp([fullfile('.',out_folder,F(ii).name) ':'])
    S = load(fullfile(F(ii).folder,F(ii).name))
end

./Collated/Dot_DataSynced_ID001_001.mat:

S = struct with fields:
    LShank: [1918x11 table]
    Pelvis: [1918x11 table]
    RShank: [1918x11 table]

./Collated/Dot_DataSynced_ID001_007.mat:

S = struct with fields:
    LShank: [1918x11 table]
    RShank: [1918x11 table]
    RThigh: [1918x11 table]

./Collated/Dot_DataSynced_ID001_011.mat:

S = struct with fields:
    LThigh: [1918x11 table]
    Pelvis: [1918x11 table]
      baba: [1918x11 table]

./Collated/Dot_DataSynced_ID001_015.mat:

S = struct with fields:
    LLCOOLJ: [1918x11 table]
     RCRUMB: [1918x11 table]
    richard: [1918x11 table]
        sal: [1918x11 table]