Skip to content

Latest commit

 

History

History
85 lines (61 loc) · 3.92 KB

changelog.md

File metadata and controls

85 lines (61 loc) · 3.92 KB
Raw

Changelog

Purpose

This file documents a summary of the changes made to the Supplementary Source Codes of the publication:

Legaspi, Roberto, & Toyoizumi, Toyoizumi (2019).
A Bayesian psychophysics model of sense of agency.
Nature Communications, 10(1), 4250.
https://doi.org/10.1038/s41467-019-12170-0

The supplementary was retrieved on 7 April 2022 from https://static-content.springer.com/esm/art%3A10.1038%2Fs41467-019-12170-0/MediaObjects/41467_2019_12170_MOESM1_ESM.pdf.

The paper and the unchanged supplementary PDF file is available in folder publication.

The overarching intent of this repository is to provide executable code from the printed supplementary.

For an exhaustive set of changes, you can inspect the Git log and compare Git commits.

Licenses

Please refer to folder LICENSES.

Summary of changes

Change summary for branch transcript

The purpose of branch transcript is to facilitate a direct visual comparison between the printed supplementary and the Matlab files in this repository.

The initial step was to use optical character recognition for bulk conversion of the printout. The conversion script and its raw output is available in folder ocr-conversion.

Subsequently, the following manual changes were made:

  • Separation of text into Matlab files, as described in the printout.

  • Matching the Matlab code with the printout by visual comparison.

    • There are two minor exceptions:
      1. Some comments are indented less than in the printout, to facilitate side-by-side comparison.
      2. Renamed all instances of function name soa_plotPrcShts.m with soa_plotPrcShfts.m for internal consistency; e.g. compare page 1 and 18 of the printout.

  • Note that this branch requires the Statistics and Machine Learning Toolbox.

Change summary for branch main

The purpose of branch main is to provide a slightly "more convenient" version of the code. None of the changes are intended to change the simulation results. Below is a summary of the changes.

A few semantic changes aim to increase code clarity and flexibility:

  • Add flag to choose whether to use Statistics and Machine Learning Toolbox.
  • Remove code repetitions of semantically unconditional cases.
  • Replace wrapper function soa_InitMatrix() with zeros() and delete the file.
  • Add references to paper equations in the comments.
  • Add function soa_Sigmoid() to match expression in Methods.
  • Add axis labels as in paper and use subplots.
  • Add functions soa_plotTauSamples() and sameaxes() to plot samples.
  • Change and add headers to clarify licensing (make REUSE compliant).
  • CCE computation: Drop a for-loop and refactor computation of terms.
  • Add function soa_IBoptimalPXi1.m to avoid code duplication.
  • Unify the three lookup functions soa_IB*.m with more robust switch/case and paper references.
  • Move repeated code for CCE into function soa_computeCCE.m and for perceptual temporal shifts into function soa_computePrcShft.m.
  • Generalize function soa_sortMatrices.m to other than 3 conditions.

The rest are opinionated, cosmetic changes to subjectively increase readability and to avoid code repetitions.

  • More consistent use of whitespaces; e.g. add spaces before commas and around inline operators, etc.
  • Add newlines and reorder to get a single instruction per line.
  • Limit line width to 80 characters.
  • Mark all function calls with parentheses ().
  • End functions with end.
  • Fix typos in comments.
  • Appease Matlab's linter.
  • Replace multi-line comments with single-line comments.
  • Replace tao with tau if referring to the Greek letter, to avoid confusion with A/O for action/outcome.
  • Add explanatory comments and paper references.
  • Prettify all printouts to the console.
  • Add helper functions soa_getFileNames.m and soa_loadTauSamples.m for sample file handling.