Count number of unique .mat files in a folder

Hello,

I have a set of folders and subfolder which I wish to analyze. I used the following code to determine the .mat files Info and names in a specific subfolder:

function [dirinfo, folders, matFilesInfo, matFilesNames] = getFolderNames( rootPath )

dirinfo = dir(rootPath);

dirinfo(~[dirinfo.isdir]) = []; %remove non-directories

tf = ismember( {dirinfo.name}, {'.', '..'});

dirinfo(tf) = []; %remove current and parent directory.

if ~isempty(dirinfo)

folders = extractfield(dirinfo,'name');

else

folders = [];

end

matFilesInfo = dir(fullfile(rootPath, '*.mat'));

if ~isempty(matFilesInfo)

matFilesNames = extractfield(matFilesInfo,'name');

else

matFilesNames = [];

end

end

For example I have the following values in matFilesNames:

6×1 cell array

{'RX_Phase_SA_RXG0_TXG20_1.mat'}

{'RX_Phase_SA_RXG0_TXG20_2.mat'}

{'RX_Phase_SA_RXG0_TXG20_3.mat'}

{'RX_Phase_SA_RXG0_TXG30_1.mat'}

{'RX_Phase_SA_RXG0_TXG30_2.mat'}

{'RX_Phase_SA_RXG0_TXG40_1.mat'}

I would like to have unique values with the count for each unique case.

For the above example:

3 files for 'RX_Phase_SA_RXG0_TXG20', 2 files for 'RX_Phase_SA_RXG0_TXG30' and 1 file for 'RX_Phase_SA_RXG0_TXG40'

How can I do this w/o huge number of loops and string splits?

Thank you!

[Stephen23]

Try this

S = dir(fullfile(rootPath,'**','*.mat'))

Answer 1

[Voss]

Your code appears to be collecting information (dirinfo, folders) about sub-directories within the given rootPath, and then also information about the mat-files within rootPath, as opposed to mat-files within the sub-directories of rootPath. I don't know whether this is what is intended.

Regardless, you have a cell array and you want to parse each element of it (assuming the relevant part to compare is the part of each element before the last underscore in it), create a set of unique elements (with just the part before the last underscore), and count how many times each one is in the cell array. Here is one way you can do that:

matFilesNames = { ...
    'RX_Phase_SA_RXG0_TXG20_1.mat'; ...
    'RX_Phase_SA_RXG0_TXG20_2.mat'; ...
    'RX_Phase_SA_RXG0_TXG20_3.mat'; ...
    'RX_Phase_SA_RXG0_TXG30_1.mat'; ...
    'RX_Phase_SA_RXG0_TXG30_2.mat'; ...
    'RX_Phase_SA_RXG0_TXG40_1.mat'}; % using your example

% get the index of the last underscore in each name
idx = cellfun(@(x)find(x == '_',1,'last'),matFilesNames);

% store the part of each name up to but not including the last underscore
matFilesNames = arrayfun(@(x,y)x{1}(1:y),matFilesNames,idx-1,'UniformOutput',false);

% now get the unique set of names and index (j) where each one from the
% original set is in the unique set
[unique_names,~,j] = unique(matFilesNames);

% and count them up
n_unique = numel(unique_names);
counts = zeros(n_unique,1);
for i = 1:n_unique
    counts(i) = nnz(j == i);
end

display(unique_names);

unique_names = 3×1 cell array
    {'RX_Phase_SA_RXG0_TXG20'}
    {'RX_Phase_SA_RXG0_TXG30'}
    {'RX_Phase_SA_RXG0_TXG40'}

display(counts);

counts = 3×1

3 2 1

[GSM]

Thanks!

Answer 2

[Stephen23]

The simple MATLAB approach is to use one of the histogram functions rather than a loop, e.g.:

% S = dir(fullfile(rootPath,'**','*.mat'));
% C = {S.name};
C = { ...
    'RX_Phase_SA_RXG0_TXG20_1.mat'; ...
    'RX_Phase_SA_RXG0_TXG20_2.mat'; ...
    'RX_Phase_SA_RXG0_TXG20_3.mat'; ...
    'RX_Phase_SA_RXG0_TXG30_1.mat'; ...
    'RX_Phase_SA_RXG0_TXG30_2.mat'; ...
    'RX_Phase_SA_RXG0_TXG40_1.mat'};
[U,~,X] = unique(regexprep(C,'_[^_]+$',''))

U = 3×1 cell array
    {'RX_Phase_SA_RXG0_TXG20'}
    {'RX_Phase_SA_RXG0_TXG30'}
    {'RX_Phase_SA_RXG0_TXG40'}

X = 6×1

1 1 1 2 2 3

N = histcounts(X)

N = 1×3

3 2 1

optional output display:

compose("%s:  %d",string(U),N(:))

ans = 3×1 string array
    "RX_Phase_SA_RXG0_TXG20:  3"
    "RX_Phase_SA_RXG0_TXG30:  2"
    "RX_Phase_SA_RXG0_TXG40:  1"