Mathematical Framework for the Study of Personality

View the Project on GitHub OrenForkosh/IdentityDomains


This code provides tools for infering personality traits from behavior. While personality traits offer considerable insight into the biological basis of individual differences, existing approaches toward understanding personality across species rely on subjective criteria and limited sets of behavioral readouts, resulting in noisy and often inconsistent outcomes. In a recent paper, we introduced a mathematical framework for studying individual differences along dimensions with maximum consistency and discriminative power. This framework was validated in mice, using data from a system for high-throughput longitudinal monitoring of group-housed male mice that yields a variety of readouts from all across an individual’s behavioral repertoire. The code is implemented in Matlab.

Getting Started

The file IdentityDemos is a good starting point as it includes several examples for using the code. The script shows how to:

  1. Compute the identity-domains and showing their behavioral correlates
  2. Use pre-computed Identity-Domains to produce a map of the personality space


The code was tested on Matlab 2017a on both a Windows and MacOS machines, but should in principal work on earlier version of Matlab as well as other operating systems


Just put all the code in a directory and cd to this directory within Matlab


File Description
IdentityDemos.m Includes a set of examples for using the code
IdentityDomains.m Main class for computing the IDs
DimReduction.m A collection of Dimensionality Reduction algorithms
Auxiliary.m Auxiliary functions
Color.m A collection of colors and tools for colors
data_table.mat Sample table
ID.mat Precomputed IDs and Archetypes


Forkosh, Oren, Stoyo Karamihalev, Simone Roeh, Uri Alon, Sergey Anpilov, Chadi Touma, Markus Nussbaumer et al. “Identity domains capture individual differences from across the behavioral repertoire.” Nature Neuroscience 22, no. 12 (2019): 2023-2028. doi:


The Identity Domains Toolbox for MATLAB is free and open source, distributed under the MIT license.