Why is the RMSE returned by the Linear Fit deviating the sqrt(sum(residuals^2)/n)?
I am confused regarding the Root Mean Square Error reported in the "Summary of Fit" in the Linear Fit platform. x = [1.309, 1.471, 1.49, 1.565, 1.611, 1.68];
y = [2.138, 3.421, 3.597, 4.34, 4.882, 5.66];
{Estimates, Std_Error, Diagnostics} = Linear Regression( y, x, <<printToLog );
z=Estimates[1]+Estimates[2]*x;
rmse_lin_reg=sqrt(sum((z-y)^2)/nrows(y));
as table(x,y,<<Column Names({"x","y"}));
b...