Prest is an open-source and user-friendly Windows & macOS desktop application.

It can be used to analyze choice datasets created by economists, psychologists and marketing researchers.

Its key novelties pertain to general datasets where choice alternatives are discrete.

Prest allows for estimating non-parametrically the decision maker’s preferences from such general datasets.

It does so by finding out how “close” the observed choices are to being explainable by rational choice or some model of bounded-rational choice.

In this way, Prest recovers both the individual’s decision rule and their preferences conditional on that rule.


Prest is open-source software and its latest version will always be available online for free.

Prest does not collect any data entered by its users.


  • Prest v1.1.0 for Windows — No installation required: run by double-clicking the .exe file.

  • Prest v1.1.0 for macOS — No installation required: run by double-clicking the .command file. Select “Open anyway” if prompted. If the “Open anyway” button is not available, close the dialog window and double-click the .command file again.

  • The Prest source code, written in Rust (core) and Python (graphical user interface).

  • New feature since Prest 1.1: visualization of preference-estimation output using GraphViz

  • Previous downloadable versions of Prest are available in the archive.


The pages linked below (and also in the navigation menu on the left) contain information about Prest’s features, define the terms used in the graphical user interface, and explain relevant background concepts.


Text boxes with the Tip label provide essential information about Prest’s features.


Text boxes with the Note label provide supplementary information.

Prest Developers

Georgios Gerasimou & Matúš Tejiščák

If you use Prest in your work, please cite it as follows:

Georgios Gerasimou and Matúš Tejiščák (2018) “Prest: Open-Source Software for Computational Revealed Preference Analysis”, Journal of Open Source Software, 3(30), 1015, doi:10.21105.joss.01015.