Andrew Reid PhD

This is the Python code corresponding to blog post 21, discussing causal discovery approaches. This code is a simplistic implementation of the PC algorithm. It can be used to (1) generate samples from the source directed acyclic graph (DAG) with associated (conditional) probabilities; and (2) attempt to discovery this DAG.

Extract the archive. Run generate_*.py, process_*.py, or test_*.py using the python command. See the README file for more details. The software was developed using Anaconda 2022.10, Python 3.9. Environment files are provided, including environment-history.yml and environment-full.yml.

This is the full ModelGUI application. It contains executables and all dependencies. For more information, see the wiki site. See Github link for source code. See this link for previous versions of ModelGUI.

Extract the archive. To run the application, execute ./mgui-amd64 or ./mgui-i586 for Linux, depending on whether your system is 64- or 32-bit. Use mgui-amd64.bat or mgui-i586.bat for Windows. For Mac, use ./mgui-mac. You must have Java JDK 10 or later installed (this is typically the case for newer systems; just trying running it first -- if you get an error you may need to download the linked JDK). More details are in the README file.

This is the full ModelGUI application. It contains executables and all dependencies. For more information, see the wiki site. See Github link for source code. See this link for previous versions of ModelGUI.

Extract the archive. To run the application, execute ./mgui-amd64 or ./mgui-i586 for Linux, depending on whether your system is 64- or 32-bit. Use mgui-amd64.bat or mgui-i586.bat for Windows. For Mac, use ./mgui-mac. You must have Java JDK 8 or later installed (this is typically the case for newer systems; just trying running it first -- if you get an error you may need to download the linked JDK). More details are in the README file.

Matlab script to generate plots demonstrating how Activation Likelihood Estimation (ALE) – a meta-analytic neuroimaging approach – works. See this article for a description of the approach. Use the file gaussians.m to run the demo.

Extract the scripts somewhere on your hard disk; done.

Simple simulation used to explore functional covariance and its relationship to physical connectivity (Blog #11).

Extract the script somewhere on your hard disk; done.

A collection of custom function scripts that are used in most of my Matlab code.

Extract the archive somewhere on your hard disk, and add the directory name to your Matlab path using [addpath](http://www.mathworks.com/help/matlab/ref/addpath.html) [path/to/directory].

This is the full ModelGUI application. It contains executables and all dependencies. For more information, see the wiki site. See Github link for source code. See this link for previous versions of ModelGUI.

Extract the archive. To run the application, execute ./mgui-amd64 or ./mgui-i586 for Linux, depending on whether your system is 64- or 32-bit. Use mgui-amd64.bat or mgui-i586.bat for Windows. For Mac, use ./mgui-mac. You must have Java JDK 8 or later installed (this is typically the case for newer systems; just trying running it first -- if you get an error you may need to download the linked JDK). More details are in the README file.

This implements a very simple 1-dimensional simulated annealing optimization, as shown in this blog entry. This is not useful for anything but didactic purposes! There are many better simulated annealing libraries out there.

Extract the archive, and use the file examples.m to run the examples. Remember to add the dependencies to your Matlab path!

This extends the 1-dimensional simulated annealing example to 2 dimensions, which for example can demonstrate some basic principles of foraging optimization. As with the 1D, this is not useful for anything but didactic purposes! There are many better simulated annealing libraries out there.

Extract the archive, and use the file forage_example.m to run the examples. Remember to add the dependencies to your Matlab path!