SEEMPS is the second iteration of the SElf-Explaining Matrix-Product-State library.
The original library, still available here was a collection of Jupyter notebooks with a well documented implementation of matrix-product state algorithms.
The current iteration aims to be more useable and have better and more standard documentation, while preserving the same degree of accessibility of the algorithms.
The library is thought out as introduction to the world of Matrix Product States and DMRG-inspired algorithms. Its main goal is not performance, but rapid prototyping and testing of ideas, providing a good playground before dwelling in more advanced (C++, Julia) versions of the algorithms.
This said, the library as it stands has been used in some heavy-duty simulations involving tens and hundreds of qubits, and, in particular, its current iteration arises from two works on quantum-inspired algorithms for numerical analysis:
-
Quantum-inspired algorithms for multivariate analysis: from interpolation to partial differential equations, Juan José García-Ripoll, Quantum 5, 431 (2021), https://doi.org/10.22331/q-2021-04-15-431
-
Global optimization of MPS in quantum-inspired numerical analysis, Paula García-Molina, Luca Tagliacozzo, Juan José García-Ripoll, https://arxiv.org/abs/2303.09430
The library is developed in a mixture of Python 3 and Cython, with the support of Numpy, Scipy and h5py. Installation instructions are provided in the documentation.
Version: 2.1
Authors:
- Juan José García Ripoll (Institute of Fundamental Physics)
- Paula García Molina (Institute of Fundmental Physics)