An overview of Gauss, to be read first, is provided in:
The Gauss code given below is applied to data on takeover bids which you need to retrieve from: The simplest Gauss programs (using dataset jaggia.asc) and associated output are in
More advanced Gauss programs use the Gauss add-on maxlik (which is a separate additional purchase). In addition to the particular program below, you need seldata.src which is called by the programs. For MLE you also need the log-likelihood, gradient and hessian. These are given in ghcount.src (for Poisson) and ghmnl.src (for logit and three-choice multinomial logit).
For example to run the Poisson MLE, you run the program jagmle2.prg, and need to have in the same directory seldata.src and ghcount.src (these two files containing procedures are inserted in jagmle2.prg by a #include statement) and dataset jaggia.asc. Poisson regression follows:
Binary logit and three-choice multinomial logit follow. For example to run the binary logit MLE, you run the program jagmnl1.prg, and need to have in the same directory seldata.src and ghmnl.src (these two files containing procedures are inserted in jagmnl1.prg by a #include statement) and dataset jaggia.asc.
The following Gauss code data on takeover bids which you need to retrieve from:
Linear fixed effects panel data follows. To run this program you need to have in the same directory seldata.src (inserted in hhgpan1.prg with an #include statement) and the dataset patr7079.dat. A. Colin Cameron / UC-Davis Economics / http://www.econ.ucdavis.edu/faculty/cameron