index out of bounds because... can't solve error

I am performing a stepwise regression and looping through data, this performs 6 stepwise regressions. I then try to output some of the data. As you can see on the code below, I want to extract the INMODEL data and SSTotal data.

When I run the code I get the outputs for INMODEL for regressions 1 and 2, fantastic. However the SSTotal only comes out for the first regression, giving me the following error for the second regression

"Attempted to access STATS.SStotal(2);

index out of bounds because

numel(STATS.SStotal)=1".

I don't understand the error, as it works for the INMODEL but not SSTotal. Any help greatly appreciated

[rows, cols]=size(A11);

        % initialize matrix to store the data
        pvalues=zeros(rows, cols)-1;
        coefficients=zeros(rows,cols);
        r2=zeros(rows,cols);

    for i = 1:rows
        for j = 1:cols
            y=zeros(8,1)
            x=zeros(8,6)
            for m=1:8
                y(m)=TOP((m-1)*rows+i,j);
                x(m,1)=Approx1((m-1)*rows+i,j);
                x(m,2)=Approx2((m-1)*rows+i,j);
                x(m,3)=Approx3((m-1)*rows+i,j);
                x(m,4)=Approx4((m-1)*rows+i,j);
                x(m,5)=Approx5((m-1)*rows+i,j);
                x(m,6)=Approx6((m-1)*rows+i,j);
            end
            %Run the stepwise Regression
            if  sum(sum(isnan(x)),2)==0 && sum(isnan(y))==0 
            xx=x(:,1:6); yy=y(:,1);

            [B,SE,PVAL,INMODEL,STATS,NEXTSTEP,HISTORY]= ...
                                           stepwisefit(xx,yy,'penter',.05);

          inApprox1(i,j)=INMODEL(1); %comes out fine
          inApprox2(i,j)=INMODEL(2); %Comes out fine

sstotApprox1(i,j)=STATS.SStotal(1); %Comes out fine
sstotApprox2(i,j)=STATS.SStotal(2); %ERROR

         end
      end
  end

Answer 1

[Walter Roberson]

The SStotal is documented as "Total sum of squares of the response". As it is a total, I would expect there to be only a single value. Perhaps you want SSresid "Sum of squares of the residuals", which possibly has one entry for each residual (but I am not certain of that.)

[Robert]

Thank you for your response, I think you're right. I get the same error for residuals. The reason I am trying to get SStotal and SSresidual is the calculate R squared, which is not a stepwisefit output, perhaps I am going about it the wrong way...