This package contains the code for reproducing the simulations of our paper:
Isabell Lehmann, B. Gabrielson, and T. Adali, Revisiting the Connection between mCCA-genvar and IVA-G: Role of Orthogonality and Deflation, submitted for ICASSP 2026.
The only pre-requisite is to have Python 3 (version >=3.11) installed. This package can be installed (optionally in a virtual environment) with:
git clone https://github.com/SSTGroup/connection_mcca_genvar_iva_g
cd connection_mcca_genvar_iva_g
pip install -e .
Required third party packages will automatically be installed.
IMPORTANT NOTE: A folder called simulation_results must manually be created in connection_mcca_genvar_iva_g before starting the simulations.
The simulations can be run by calling the functions test_save_paper_results_alpha() and test_save_paper_results_samples() in test_experiments.py.
After running the code, the folder connection_mcca_genvar_iva_g/simulation_results will contain the generated .npy files.
The generated SCV covariance matrices can be visualized by calling the function test_plot_scv_covs_one_run() in test_experiments.py.
The results can be plotted by calling the functions test_plot_results_for_different_alpha() and test_plot_results_for_different_samples() in test_experiments.py.
In case of questions, suggestions, problems etc. please send an email to isabell.lehmann@sst.upb.de, or open an issue here on Github.
If you use this code in an academic paper, please cite [1]
@inproceedings{Lehmann2026,
title = {Revisiting the Connection between mCCA-genvar and IVA-G: Role of Orthogonality and Deflation},
author = {Lehmann, Isabell and Gabrielson, Ben and Adali, T{\"u}lay},
booktitle={submitted for review},
year={2026}
}
[1] Isabell Lehmann, B. Gabrielson, and T. Adali, Revisiting the Connection between mCCA-genvar and IVA-G: Role of Orthogonality and Deflation, submitted for review, 2026.