No-frills, SciPy-based implementation of the regularized Laplacian kernel for smoothing node scores across a network.
deborah-chasman / regularized-laplacian-kernel-py Goto Github PK
View Code? Open in Web Editor NEWNo-frills, SciPy-based implementation of the regularized Laplacian kernel for smoothing node scores across an undirected network. Works easily with graphs stored in "sif" format (Cytoscape friendly) or tab-delimited format. Doesn't do anything smart about inverting the matrix, but it does seem to work fine on ~100k-node matrices given enough memory.
License: GNU General Public License v2.0