Here, we demonstrate functional alignment methods using simulated data as well as the publicly accessible Learning Naturalistic Structure dataset, generously shared by Aly and colleagues. For more information on the acquisition of this data set, please see their paper:
Aly M, Chen J, Turk-Browne NB, & Hasson U (2018). Learning naturalistic temporal structure in the posterior medial network. Journal of Cognitive Neuroscience, 30(9): 1345-1365.
Running on Binder
Unfortunately, several of these tutorials are too computationally intensive to run on the public mybinder instance. Although we are exploring alternative binderhub instances, for now only the simulation tutorial is available to run in-browser.
To access this material locally, you can download this repository and install the requirements using
We recommend that you install these requirements within a virtual environment;
for example, the following commands will create a conda environment and install all necessary packages:
git clone https://github.com/neurodatascience/fmralign-tutorials cd fmralign-tutorials conda create --name fmralign-tutorials python=3.6 source activate fmralign-tutorials pip install -r requirements.txt
You can render the tutorials using
which provides a convenient means to sync the user-friendly notebook interface with a git-friendly plain-text Python script.
For example, for the
aly_benchmark tutorial, simply run:
jupytext aly_benchmark.py --to notebook jupytext --sync aly_benchmark.py aly_benchmark.ipynb
You can then launch the notebook using:
jupyter notebook aly_benchmark.ipynb
All changes you make there will be automatically updated in the associated Python script.