GAUSS Programs for

DISCRETE CHOICE MODELLING IN AIRLINE NETWORK MANAGEMENT
=======================================================

by Joachim Grammig, Michael Scheidler and Reinhard Hujer

forthcoming Journal of Applied Econometrics

the programs produce the results of
forthcoming version

http://www.uni-tuebingen.de/uni/wwo/Grammig/jae_ghs2005_final.pdf


Contact:
author
{{\bf Joachim Grammig}${}^1$ \\ University of T\"{u}bingen\\
Department of Statistics and Econometrics\\ Mohlstr. 36, D-72074
T\"{u}bingen, Germany \\ Ph: +49 7071 29 78159. Fax: +49 7071 29
5546\\ e-mail: joachim.grammig@uni-tuebingen.de \\ {\bf Reinhard
Hujer}\\ University of Frankfurt \\ Faculty of Economics and
Business Administration\\ Mertonstr. 17, D-60054 Frankfurt,
Germany \\ {\bf Michael Scheidler}\\ University of Frankfurt \\
Faculty of Economics and Business Administration \\ Mertonstr. 17,
D-60054 Frankfurt, Germany }

The programs can be freely used for academic purposes if
this source is acknowledged.

To run the programs copy the contents of the zip file in a
directory.

1. To reproduce the estimation results
======================================

The GAUSS programs ("front end programs")

d_estnl.prg
d_estmnp.prg
d_estmnl.prg
d_estgar.prg
d_estabc.prg

are used to produce the estimation results in table II. They perform
ML estimation of the respective model.

Note that the MNP models are time consuming. You have to be patient
on a slow computer. Try a MNL or a NL first.

These "front end" programs
#include a set of auxiliary procedure files, namely

MY_MNP.prg
MY_gar.prg
MY_MNL.prg
DChoice0.prg
DChoiceN.prg
DChPrgs.prg
DChUti_A.prg
DChUti_C.prg
DChUti_D.prg
DChUti_E.prg
DNLPrgs.prg
DNLUti_A.prg
LKE_MNP.PRG

These programs are nested, i.e. some of these procedure files
#include others. These procedure files include the procedures to
evaluate the likelihood functions of the respective models compute tests et cetera.

The data used is in the files
estdata
available as ASCII (estdata.asc) EXCEL (estdata.xls) and Gauss
(estdata.dat/dht)

2. To run the simulations for the inserted flight (figure 2)
===========================================================
s_simabc.prg
s_simgar.prg
s_simmnl.prg
s_simmnl.prg

These programs use a data set called markt1ad which is available
in GAUSS format and in ASCII format

Compatibility and requirements:
===============================
1. You need the GAUSS modules CML (Constrained ML) and GAUSSX
installed on your machines.

2. The results were produced using GAUSS 3.2.3. We tested the programs
under GAUSS 6.0 and those tested run without problems. The programs used to
simulate the effect of an inserted flight using the GAR and the ABC MNP
do not run under GAUSS 6.0. The problem is with the GAUSSX qdfn.src procedure used
to calculate multivariate normal probabilities.
In order to run the program under GAUSS 6.0 you have to change the procedure
_qdfdas in qdfn.src by making the (global) variable _intrec3 a local variable.
