Error -Index exceeds matrix dimensions.

I get error ,please tell how to rectify it,i have used 100x4 for training and rest for testing, even though i get error

clc
clear all
load fisheriris
groups = ismember(species,'setosa');
k=10;
groups =groups(1:100,:);
mt=meas(1:100,:);
mte=meas(101:end,:);
cvFolds = crossvalind('Kfold', groups, k);
cp = classperf(groups);
for i = 1:k
    testIdx = (cvFolds == i);
    trainIdx = ~testIdx;

      svmModel = svmtrain(mt(trainIdx,:), groups(trainIdx), ...
          'Autoscale',true, 'Showplot',false, 'Method','QP', ...
          'BoxConstraint',2e-1, 'Kernel_Function','rbf', 'RBF_Sigma',1);

      pred = svmclassify(svmModel, mte(testIdx,:), 'Showplot',false);

      cp = classperf(cp, pred, testIdx);
  end

[Walter Roberson]

Usually we can debug faster when the person tells us which line the error is occuring on.

[nkumar]

 Index exceeds matrix dimensions.

Error in ==> Untitl at 22 pred = svmclassify(svmModel, mte(testIdx,:), 'Showplot',false);

[Jan]

Btw., Don't clear all.

Answer 1

[Walter Roberson]

Your cvFolds is constructed from groups, which is rows 1:100, but you use an index derived from that (testIdx) to index mte which is rows 101:end and so might be smaller.

[nkumar]

So how to rectify it walter

[Jan]

How could we know, what you exactly want to achieve?

[nkumar]

i want to train the data with 100 data ant test the remaining