type of variable "m" changes (i.e int or [ ] ) while running a program using different datasets

when i use a dataset : values of m,n,temp are integer and, i get results

but, when i use another dataset as input : value of m becomes an empty array [] , i cant figure out how ??

here the code

rankloss=0;

    for i=1:num_instance
        temp=0;
        for m=1:Label_size(i)
            for n=1:(num_class-Label_size(i))
                if(Outputs(Label{i,1}(m),i)<=Outputs(not_Label{i,1}(n),i))
                    temp=temp+1;
                end
            end
        end
         *rl_binary(i)=temp/(m*n); 
        rankloss=rankloss+temp/(m*n);* 
    end

 Label_size=zeros(1,num_instance); and 

 [num_class,num_instance]=size(Outputs); here.

so as a result i get this error

Error using / : Matrix dimensions must agree: temp/(m*n); (because m is [] )

i tried temp./(m.*n) but it obviously gives an empty array [] as a reult

note: both input datasets have same dimensions and same type , only data differs

i hoep my question is clear .... how can i figure out why m becomes [ ] instead of an integer value ??

Answer 1

[Chad Greene]

Is Label_size defined before this bit of code?

[pooja]

yes Mr.Chad Greene i have specified in my question (sorry 4 that gray text in it )

    Label_size=zeros(1,num_instance); and 

   [num_class,num_instance]=size(Outputs); here.