This repository includes the analysis as described in Emergence of Canonical Functional Networks from Complex-valued Laplacian of Structural Connectome, where we investigate whether we can recreate the canonical functional networks of the human brain from structural connectivity with a simple low dimensional framework.

This repository is dependent on spectrome, please see instructions in the repository for setting up your conda environment. The spectrome repository will need to be cloned, and the path to the spectrome folder will need to be appended to $PYTHONPATH. Activating the spectrome environment will enable all the analysis in this repository.

If you wish to visualize brain renderings, you will need to add ipywidgets and pysurfer dependencies to your environment. The current public spectrome environment is made for Binder, meaning we removed any pop out visualization tools that caused issues with Binder.

• notebooks: Jupyter notebooks that produced the paper results.
• scripts: python script for optimization.
• data: some intermediate data needed to generate the figures.
  • mean80_fiber*.csv: HCP averaged fibercount and fiberlength files.
  • DK_dictionary_normalized.csv: The canonical functional networks in DK-atlas parcellations