A developing project for simulating dynamics of 1D quantum systems using tensor networks.
The main project folder is named hellomps/ and is organized as follows:
networks/ contains commonly used tensor network architectures, MPS for pure states and LPTN for mixed states. operations.py integrates a variety of tensor networks operations, which are essential to the simulation algorithms. mpo_projected.py offers the abstract interface between projected MPO and iterative solvers, which is the central unit to variational optimization problems, such as DMRG.
models/ provides models used for simulation, such as transverse field Ising model, Bose-Hubburd model, etc. This part is still quite basic as for now.
algorithms/ implements ground state serach algorithms for MPS, real and imaginary time evolution algorithms for both MPS and LPTN.
- Sebastian Paeckel, Thomas Köhler, Andreas Swoboda, Salvatore R. Manmana, Ulrich Schollwöck, Claudius Hubig, Time-evolution methods for matrix-product states Annals of Physics, Volume 411, 2019
- Werner, A. H. and Jaschke, D. and Silvi, P. and Kliesch, M. and Calarco, T. and Eisert, J. and Montangero, S. Positive Tensor Network Approach for Simulating Open Quantum Many-Body Systems Phys. Rev. Lett. 116, 237201
- Jutho Haegeman, Christian Lubich, Ivan Oseledets, Bart Vandereycken, and Frank Verstraete Unifying time evolution and optimization with matrix product states Phys. Rev. B 94, 165116