Skip to content

Simulate an experiment with optimal stimuli segmentation

Notifications You must be signed in to change notification settings

msieg/OptSeg_Reproducible

Repository files navigation

OptSeg_Reproducible

Description

OptSeg is a tool for optimally segmenting stimuli into matched lists. This GitHub Repo contains a markdown file OptSeg.md that allows you to simulate a neuroscience experiment with and without OptSeg. The output of OptSeg.md is OptSeg.html, a fully reproducible report that details the results of the experiment.

How to run

Simply clone this repo and execute the bash script run_experiment.sh from the command line to install the python package requirements and simulate the neuroscience experiment.

e.g.

git clone https://github.com/msieg/OptSeg_Reproducible.git

cd OptSeg_Reproducible

bash run_experiment.sh

About

Simulate an experiment with optimal stimuli segmentation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages