MO|DE.behave is a Python-based software package for the estimation and simulation of discrete choice models and shall enable rapid quantitative analyses based on survey data on choice behavior, utilizing advanced discrete choice methods. Please refer to ''.


What MO|DE.behave can do for you

MODE.behave incorporates estimation routines for conventional multinomial logit models, as well as for mixed logit models with nonparametric distributions. Furthermore, MO|DE.behave contains a set of post-processing tools for visualizing estimation and simulation results. Additionally, pre-estimated discrete choice simulation methods for transportation research are included to enrich the software package for this specific community.

On mixed logit models: In recent years, a new modeling approach in the field of discrete choice theory became popular – the mixed logit model (see Train, K. (2009): "Mixed logit", in Discrete choice methods with simulation (pp. 76–93), Cambridge University Press). Conventional discrete choice models only have a limited capability to describe the heterogeneity of choice preferences within a base population, i.e., the divergent choice behavior of different individuals or consumer groups can only be studied to a limited degree. Mixed logit models overcome this deficiency and allow for the analysis of preference distributions across base populations.

Communication and contribution: We encourage active participation in the software development process to adapt it to user needs. If you would like to contribute to the project or report any bugs, please refer to the contribution-file or simply create an issue in the repository. For any other interests (e.g. potential research collaborations), please directly contact the project maintainers via email, as indicated and updated on GitHub.

Documentation on GitHub Pages:

No keywords available
Programming languages
  • Python 98%
  • TeX 2%
</>Source code

Participating organisations

Forschungszentrum Jülich