Re: [Gretl-users] Weighted least squares

I seem to be missunderstanding something about WLS in gretl. My understading is the WLS option from the dropdown model Menu weighs the variables with the weights series. So if I have a regression of y on a constant and x and I add a weight series w in the WLS menu, then the regression will be y*w on 1*w (the constant) and x*w. However, if I do this manually by constricting variables y*w, 1*w, x*w and then run an OLS regression on these variables I get different parameter estimates. What am I missing? _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users