CGN_pipelines

Quantifying similarities between fMRI processing pipelines for efficient multiverse analysis

Installation

As stellargraph requires Python 3.8, we recommend using a dedicated environment (e.g. through conda):

conda create -n gcn-piplines python==3.8.0
conda activate gcn-piplines
pip install -r requirements.txt

To re-run the multiverse analysis, a separate environment with Python 3.11 is required:

conda create -n gcn-multiverse python==3.11
conda activate gcn-multiverse
pip install -r requirements_multiverse.txt

The main script which reproduces the results fom the paper is located in evaluate_gcn.ipynb
The multiverse analysis can be performed by running multiverse/run_multiverse.py
The results from multiverse analysis are visualized in evaluate_multiverse.ipynb

Name		Name	Last commit message	Last commit date
Latest commit History 16 Commits
data		data
figures		figures
multiverse		multiverse
1_create_embeddings.ipynb		1_create_embeddings.ipynb
2_evaluate_gcn.ipynb		2_evaluate_gcn.ipynb
3_evaulate_multiverse.ipynb		3_evaulate_multiverse.ipynb
4_create_paper_results.ipynb		4_create_paper_results.ipynb
LICENSE		LICENSE
README.md		README.md
evaluate_gcn.ipynb		evaluate_gcn.ipynb
evaulate_multiverse.ipynb		evaulate_multiverse.ipynb
prepare_data.ipynb		prepare_data.ipynb
requirements.txt		requirements.txt
requirements_multiverse.txt		requirements_multiverse.txt