Re: [Gretl-users] Nonlinear switching regression vs Loop constructs

First of all I would like to express most sincere thanks to everybody who answered on the matter. All gretl users must have seen this very fine result of scripting and it’s rare to receive fast and quality help and get exactly what you were looking for when dealing with econometric packages. Allin and Jack – I may become your fan. I have made a few modifications to the script such as: 1. Cost was scaled the same way as Price (it’s measured in the same units in the dataset) 2. Possible values of mu and sigma changed and the results are still quite consistent with the conventional theory .18<mu<.5 .01<sig<.041 3. The line for optimal mu and sigma remain and non-optimal deleted in order to avoid confusion A question: why estimate ‘y – GDV’ and not ‘y’ as a dependent variable. I understand that it’s equivalent in the linear regression just want to know the reason for this. The problem with the script is - NLS isn’t estimated if FE = 1 because series u = $unit doesn’t work with the current periodicity. Some pointers on how to solve this would be nice. Another problem I encountered is my little knowledge of the NLS estimation process – I can’t figure out how Jack uses linear combinations, nelem function and covarince matrix to get the NLS results. If somebody could recommend me some scientific articles or a good advanced textbook that covers the technique similar to the one used in the script it would definitely help. Regards, Gregory On Sat, Jan 29, 2011 at 6:33 PM, Riccardo (Jack) Lucchetti <r.lucchetti(a)univpm.it&gt; wrote: