The package Biogeme (biogeme.epfl.ch) is designed to estimate the pa-rameters of various models using maximum likelihood estimation. It is par-ticularly designed for discrete choice models. In this document, we presentstep by step how to specify a simple model, estimate its parameters andinterpret the output of the package. We assume that the reader is alreadyfamiliar with discrete choice models (Ben-Akiva and Lerman,1985), andhas successfully installed PandasBiogeme. Note that PythonBiogeme andPandasBiogeme have a very similar syntax. The difference is that Python-Biogeme is an independent software package written in C++, and using thePython language for model specification.PandasBiogeme is a genuine Python package written in Pythonand C++,that relies on the Pandas library for the management of the data. This is thestandard mode of operations of more and more data scientists. The syntax formodel specification is almost identical, but there are slight differences, thatare highlighted at the end of the document. This document hasbeen writtenusing PandasBiogeme 3.2.6, but should remain valid for future versions.
2020
TRANSP-OR publications
NON-REVIEWED