Package for the data-driven representation of non-linear dynamics over manifolds based on a statistical distribution of local phase portrait features. Includes specific example on dynamical systems, synthetic- and real neural datasets. https://agosztolai.github.io/MARBLE/
The computation of eigenvectors in preprocessing() is extremely slow for large (10k-100k rows/cols) matrices. It is the bottleneck in the computation. It would be good to work with fewer eigenvectors in geometry.compute_eigendecomposition(). A placeholder argument is already there.
The computation of eigenvectors in preprocessing() is extremely slow for large (10k-100k rows/cols) matrices. It is the bottleneck in the computation. It would be good to work with fewer eigenvectors in geometry.compute_eigendecomposition(). A placeholder argument is already there.