Code Modules for Finding Structure in Time: Visualizing and Analyzing Behavioral Time Series

This is the github repository containing the five Matlab code modules with sample examples introduced in the paper, Finding Structure in Time: Visualizing and Analyzing Behavioral Time Series. Here is the paper（please see updated Figure 6 here). These code modules aim to facilitate behavioral researchers with varying degrees of programming skills to interpret and analyze high-density multi-modal behavioral data.

The five modules include:

1. a step-by-step “programming basics” tutorial that introduces novice programmers to common behavioral timeseries data as well as scripts necessary to import and manipulate these data. It also provides scripts for transforming data to and from common data formats used across all modules to facilitate modifying module material to work with user data,
2. scripts to visualize the raw behavioral time series,
3. scripts to describe the distributional structure of temporal events: Burstiness calculation. This is a method to quantify the temporal regularity of occurrence of events (Goh & Barabási, 2008),
4. scripts to characterize the nonlinear dynamics over multiple timescales with Cross-Recurrence Quantification Analysis (CRQA) (Zbilut, Giuliani & Webber, 1998),
5. scripts to quantify the directional relations among a set of interdependent multimodal behavioral variables with Granger Causality Granger, 1969; Bressler & Seth, 2011).

The five modules are complementary to each other, yet each module is standalone. Novice programmers will benefit most from the modules if they carefully work through the material in Modules 1 and 2 before attempting to run or modify the later modules. More experienced programmers can more selectively focus on the modules of greatest interest to them, modifying scripts to meet their own analysis goals.

System requirement and Matlab installation

Our scripts can be run on Matlab version 2018a and later versions. Matlab is available on all three main types of operating systems: Windows, macOS, and Linux. While Matlab is not free and open source which presents limitations to our users, many institutions offer Matlab for free. One can check this following website to see if a free campus license is available here. Here are the instructions for Matlab installation. Alternatively, users can install and run our code modules on GNU Octave which is open source and runs on GNU/Linux, macOS, BSD, and Windows. This page contains the installation information for GNU Octave.

For Octave users, please download statistics package from here and image package from here. After open Octave, please install and load the packages by typing the commands below:

pkg install statistics-1.4.1.tar.gz
pkg load statistics
pkg install image-2.12.0.tar.gz
pkg load image

STEPS to use these toolkits:

1. Download a clone of this toolkit in your local folder. You can click the green button Clone or download on this page to download a ZIP file. Or, if you use Git Bash, simply type in:

git clone https://github.com/findstructureintime/Time-Series-Analysis.git

2. Open Matlab and set your working path to the folder containing the downloaded toolkit.

3. Go to each module folder, view instructions in README.md/pdf and start running!