This repository contains functions and methods to analyze fMRI subject movie data. Specifically, the focus is on multivariate searchlight driven approaches for intersubject correlations and other measures through time.
###fmri/wagner_2010/ ### All of the results and scripts applicable to Dr. Dylan Wagner's 2010 study on the movie Matchstick Men
load_data.py
This is a convenience file with methods to load in the 2010 data.
analysis.py
Various adjustable tests for running on the 2010 dataset. Mostly specific for the 2010 data, but this is a good place to
see how to use some of the modules and methods for analysis
results/data/
Pickle files of data small enough to save to GitHub
results/figures/
Graphs and plots of some results
results/nifti/
NIFTI files of results
results/tutorial.ipynb Change to this directory and run `ipython notebook` to open this interactive tutorial. It should give users a headstart
on using this toolkit.
The general utility scripts with methods for pre and post processing fMRI timeseries data
fmri_preprocessing.py
Here are all of the functions for combining, splicing, and detrending data raw NIFTI data from the fMRI.
dataset_utilities.py
This module holds all of the functions to run analyses on preprocessed Datasets
measures.py
All of the methods in this file are used in searchlight analyses
Directory with NIFTI masks
Run this script from this directory to temporarily add all modules from the fmri/fmri directory into your workspace