To repair the error about classification

indices = crossvalind('Kfold',label,5);
cp=classperf(label);
for i = 1:5
 test = (indices == i); train = ~test;
 class =knnclassify(dim(test,:),dim(train,:),label(train,:),5,'euclidean', 'nearest');
 classperf(cp,class,test)
end
cp.ErrorRate

The error:

Undefined function or variable 'label'.

Error in knn (line 1)

indices = crossvalind('Kfold',label,5);

Answer 1

[Star Strider]

The ‘label’ variable must have been defined and must exist in your workspace before it is used as an argument to crossvalind.

[Frisda Sianipar]

how to make it sir?

[Frisda Sianipar]

label = ({'Nevus','Melanoma','Pigmented','Bacal'});
indices = crossvalind('Kfold',label,5);
cp=classperf(label);
for i = 1:5
 test = (indices == i); train = ~test;
 class =knnclassify(dim(test,:),dim(train,:),label(train,:),5,'euclidean', 'nearest');
 classperf(cp,class,test)
end
cp.ErrorRate

After i change the code the error is change too:

Undefined function or variable 'dim'.

Error in knn (line 6)

class

=knnclassify(dim(test,:),dim(train,:),label(train,:),5,'euclidean',

'nearest');

[Star Strider]

It wouild appear to be a new version of the same problem.

I cannot find documentation on knnclassify, since it was removed in R2017b, and replaced by fitcknn.

[Frisda Sianipar]

so i should change dim to meas?

[Star Strider]

I have no idea what either of them are, or what you are doing with them.

Use the one that works in your code, and gives the result you want.

[Frisda Sianipar]

I want to randomly share the data using the code using k-fold cross validation

[Star Strider]

I do not understand.