Eshed Margalit, Keith Jamison, Kevin Weiner, Luca Vizioli, Ru-Yuan Zhang, Kendrick Kay*, Kalanit Grill-Spector*

* Co-senior author

This repository contains code to analyze data, compute statistical test results, and make bits of figures that are combined later in third-party software.

Code relating to figures, statistical tests, and data analyses is provided in subdirectories figures/, stats/, and analyses/, respectively. Each subdirectory contains a README explaining its contents, linked here:

1. Figures README 📊
2. Statistical Analyses README 💻
3. Data Analyses README 🔬

The code in figures/ and stats/ can be run on your machine assuming the following dependencies:

python 3.6+
R (callable with Rscript from command line)

nmle
sjstats

To run the code in figures/ and stats/, you must have the submm package installed. I recommend installing into a virtual environment.

1. From the project root directory, you can install the submm package with pip install -e .
2. Modify the project root directory in submm/constants.py, in particular line 7.
3. Modify figures/make_all.sh to source the virtualenv you created, or delete the corresponding line if you installed into your system python.

If you want to create all of the figures (and then some!) you can run make_all.sh

1. The outputs in analyses/analysis_outputs/ are generated with the scripts from analyses/ and provided here to make figure generation and statistical testing easy.
2. You'll need to edit stats/params.py and figures/params.py to point the scripts to the absolute path where the analysis_outputs/ directory lives.

The code in analyses can not, in general, be run on your machine, as it depends on absolute paths to FreeSurfer surfaces and timeseries data. Please see the data availability statement in Kay et al., 2019 for more.