TeNPy Forum

What I had in mind were the eigenvectors diagonalizing \mathcal{H}_A where \rho_A \propto e^{-\mathcal{H}_A} (suppose A is a sub-region of a finite system). I suppose these are probably related to the Schmidt vectors that you've mentioned. Would one access these for an MPS via: tenpy.networks.mpo.MP...

Maybe I'm missing something, but in addition to computing the entanglement spectrum, I'm curious to look at some of the low-lying entanglement eigenvectors -- is this prohibitively expensive because it involves explicitly constructing the reduced density matrix?