/visvest-causinf

Bayesian comparison of causal inference strategies in multisensory heading perception

Primary LanguageMATLABMIT LicenseMIT

Bayesian Comparison of Explicit and Implicit Causal Inference Strategies in Multisensory Heading Perception

This repository accompanies the manuscript by Acerbi et al. (2018), published in PLoS Computational Biology [1]. It includes human subjects' data and the code used for fitting and comparing the models reported in the paper.

The repository

The original working name of the repository was VestBMS (standing for Bayesian Model Selection of Vestibular experiment), which appears as prefix of many functions and folders.

The structure of the repository is as follows:

  • Analytics: Functions to analyze data and results of model fits.
  • ModelWork: Core model fitting functions (to be used in conjunction with the ModelWork toolbox).
  • PlotFunctions: Functions for plotting results.
  • data: Raw data for each subject, in .mat form.
  • scripts: Bash scripts used to submit jobs on the computer cluster (using SLURM).
  • supplement: Files used to analyze data in de Winkel et al. (2017), for a control analysis reported in the Supporting Information.

Please contact me at luigi.acerbi@unige.ch if you have any questions.

Reference

  1. Acerbi*, L., Dokka*, K., Angelaki, D. E. & Ma, W. J. (2018). Bayesian Comparison of Explicit and Implicit Causal Inference Strategies in Multisensory Heading Perception, PLoS Computational Biology 14(7): e1006110. (*equal contribution; link)

License

This code is released under the terms of the MIT License.