Error in classperf() while training CNN

Hi,

I am trying to classify a set of ECG Data in two classes using CNN.

Iam having input of 40 records as 40x15000 and output as 40x1 categorical.

% layers defined as the CNN architecture model and % option as training option(sgdm)
           
Indices = crossvalind('Kfold',output , 10);
cp = classperf(output);
for i=1:10
     test = (Indices == i); 
     train = ~test;
net = trainNetwork(input(train,:),output(train,:),layers,options);
class = classify(net,input(test,:));
           classperf(cp,class, test);
end
cp.CountingMatrix();

for trainNetwork output must be categorical.

Error using classperf (line 206)

Ground truth must be a cell array of character vectors or a numeric array.

how to solve this error

I am using matlab2018a

please help!!

Thanks