Re: [Gretl-users] estimating discrete choice models