Computes the Procrustes disparity between two matrices. This measures how different two point configurations are after optimal translation, rotation, and scaling alignment.
Details
The algorithm follows scipy.spatial.procrustes: 1. Center both matrices (subtract mean) 2. Normalize both matrices (divide by Frobenius norm) 3. Find optimal rotation R via SVD of X'Y 4. Return sum of squared differences after alignment
Reference: Gower, J.C. (1975). Generalized Procrustes analysis. Psychometrika, 40, 33-51.