how to convert categorical data (76*1) of labels into target data of 3 classes (76*3) for nrptool?

I have assigned labels to my database like "traindatabase.Labels" which gives 76*1 categorical data (metal pipe-19, plastic box-28, steel box-29). I need to classify into 3 classes and my target data should be 76*3 which is in the form of

1 1 1....1(19) 0 0 0... (remaining)

0 0 0..0(19) 1 1 1...1(28) 0 0...0(remaining)

0 0 0......0(47) 1 1 1...1(29)

how to make target like this? please help me

path1='D:\matprog\matfiles2\trainfiles'; % training path
path2='D:\matprog\matfiles2\testfiles';  % testing path

 traindb=imageDatastore(path1,'IncludeSubfolders',true,'LabelSource','foldernames');
testdb=imageDatastore(path2,'IncludeSubfolders',true,'LabelSource','foldernames');

%% Training
img=readimage(traindb,1);
 [pixelCounts GLs] = imhist(img); % GL-gray levels
     % Get the number of pixels in the histogram.
numberOfPixels = sum(pixelCounts);
%     Get the mean gray lavel.
    meanGL = sum(GLs .* pixelCounts) / numberOfPixels
%     Get the variance, which is the second central moment.
    varianceGL = sum((GLs - meanGL) .^ 2 .* pixelCounts) / (numberOfPixels-1)
%     Get the standard deviation.
    sd = sqrt(varianceGL);
%     Get the skew.
    skew = sum((GLs - meanGL) .^ 3 .* pixelCounts) / ((numberOfPixels - 1) * sd^3)
%     Get the kurtosis.
    kurtosis = sum((GLs - meanGL) .^ 4 .* pixelCounts) / ((numberOfPixels - 1) * sd^4)
 %     Get the entropy.
     ent = entropy(img)
FeatureVector = [meanGL varianceGL sd skew kurtosis ent];
featuresize=length(FeatureVector)
totaltrainimages=numel(traindb.Files);
trainingfeatures=zeros(totaltrainimages,featuresize,'single');

for i=1:totaltrainimages  % train all the images in training folder
    img=readimage(traindb,i);
    [pixelCounts GLs] = imhist(img); % GL-gray levels
     % Get the number of pixels in the histogram.
numberOfPixels = sum(pixelCounts);
%     Get the mean gray lavel.
    meanGL = sum(GLs .* pixelCounts) / numberOfPixels;
%     Get the variance, which is the second central moment.
    varianceGL = sum((GLs - meanGL) .^ 2 .* pixelCounts) / (numberOfPixels-1);
%     Get the standard deviation.
    sd = sqrt(varianceGL);
%     Get the skew.
    skew = sum((GLs - meanGL) .^ 3 .* pixelCounts) / ((numberOfPixels - 1) * sd^3);
%     Get the kurtosis.
    kurtosis = sum((GLs - meanGL) .^ 4 .* pixelCounts) / ((numberOfPixels - 1) * sd^4);
    %     Get the entropy.
     ent = entropy(img);
FeatureVector = [meanGL varianceGL sd skew kurtosis ent];
featuresize=length(FeatureVector);
trainingfeatures(i, 1 : featuresize) = FeatureVector;
end

traininglabels=traindb.Labels; % assign labels for training

Answer 1

[Vinayak]

Hi Santhosh

As you mentioned the labels are a categorical data of 76x1. You can easily convert them into double matrix(target matrix) using the `grp2idx` function from the Statistics and Machine Learning Toolbox.

It can also work with any custom dataset:

% Create a random categorical dataset for testing 
categories = {'metal pipe', 'plastic box', 'steel box'}; 
numSamples = 5; %Testing 
labels = categorical(randi([1, 3], numSamples, 1), 1:3, categories); 
% Convert categorical labels to numerical array 
labelIndices = grp2idx(labels) 

labelIndices = 5x1

1 2 2 3 2

% Generate one-hot encoded target matrix 
targetMatrix = full(ind2vec(labelIndices'))' 

targetMatrix = 5x3

1 0 0 0 1 0 0 1 0 0 0 1 0 1 0

Refer to the documentation on `grp2idx` and `ind2vec` for more information.

• https://www.mathworks.com/help/stats/grp2idx.html
• https://www.mathworks.com/help/deeplearning/ref/ind2vec.html