error -index must be a positive integer or logical.

I have code for rbf function with cv

xdata=[0.5 25 1 0 ;8 9 21 0;12 25 1 0 ;8 9 21 0; 12 25 1 1 ;8 9 21 0]
ydata=[2 6 9 .6 ;2  8 7 4 ;2 6 9 8 ;2  8 7 4;2 6 9 8 ;2  8 7 4]
training=xdata(:,1:4);test=xdata(:,3:4);
groups=xdata(:,end);cp = classperf(groups);  
samplesize=size(xdata);
    samplesize=size(xdata);
      c=cvpartition(samplesize,'kfold',2);
      cp = classperf(groups);  
      for i=1:2
       trainInd = find(c.training(i));

         testInd = find(c.test(i));

          trainXMat=xdata(xdata(trainInd));
          tstXMat=xdata((xdata(testInd)));
          trainYMat=ydata(ydata(trainInd));
          tstYMat=ydata((ydata(testInd)));
          cp = classperf(cp, trainInd , trainInd )
        end

i get error as

Attempted to access xdata(0.5); index must be a positive integer or logical.

Error in ==> yasrbf at 20
        trainXMat=xdata(xdata(trainInd));

please help

some of my data in xdata and ydata have values in range 0 to 1

Answer 1

[Walter Roberson]

Why are you using xdata(xdata(trainInd)) instead of just xdata(trainInd) ?

[kash]

Thanks walter it works but

CorrectRate: 0

ErrorRate: 1 ,why do i get like these

Answer 2

[Jan]

As the error message tells clearly, indices must be positive integers. You cannot access the element between the zero'th and the first one like "xdata(0.5)".

Perhaps you want xdata(trainInd) without a further nesting.

[kash]

Thanks JAn it works but

CorrectRate: 0

ErrorRate: 1 ,why do i get like these