arXiv subjects

Martin P. Thompson

Publications and source records attributed to Martin P. Thompson.

3 recordsLinked to original sources

arXiv2021

DMRjulia: Tensor recipes for entanglement renormalization computations

Detailed notes on the functions included in the DMRjulia library are included here. This discussion of how to program functions for a tensor network library are intended to be a supplement to the other documentation dedicated to explaining the high level concepts. The chosen language used here is the high-level julia language that is intended to provide an introduction to provide a concise introduction and show transparently some best practices for the functions. This document is best used as a supplement to both the internal code notes and introductions to the subject to both inform the user about other functions available and also to clarify some design choices and future directions. This document presently covers the implementation of the functions in the tensor network library for dense tensors. The algorithms implemented here is the density matrix renormalization group. The document will be updated periodically with new features to include the latest developments.

quant-ph↗

arXiv2021

Build your own tensor network library: DMRjulia I. Basic library for the density matrix renormalization group

An introduction to the density matrix renormalization group is contained here, including coding examples. The focus of this code is on basic operations involved in tensor network computations, and this forms the foundation of the DMRjulia library. Algorithmic complexity, measurements from the matrix product state, convergence to the ground state, and other relevant features are also discussed. The present document covers the implementation of operations for dense tensors into the Julia language. The code can be used as an educational tool to understand how tensor network computations are done in the context of entanglement renormalization or as a template for other codes in low level languages. A comprehensive Supplemental Material is meant to be a "Numerical Recipes" style introduction to the core functions and a simple implementation of them. The code is fast enough to be used in research and can be used to make new algorithms.

quant-ph↗

arXiv2021

Méthodes de calcul avec réseaux de tenseurs en physique (Basic tensor network computations in physics)

Cet article se veut un survol des réseaux de tenseurs et s'adresse aux débutants en la matière. Nous y mettons l'accent sur les outils nécessaires à l'implémentation concrète d'algorithmes. Quatre opérations de base (remodelage, permutation d'indices, contraction et décomposition) qui sont couramment utilisées dans les algorithmes de réseaux de tenseurs y sont décrites. Y seront aussi couverts la notation diagrammatique, les états en produit de matrices (MPS), les opérateurs en produit de matrices (MPO), état projeté de paires intriquées (PEPS), l'approche par renormalisation d'enchevêtrement multi-échelle (MERA), la décimation par bloc d'évolution temporelle (TEBD) et le groupe de renormalisation de tenseurs (TRG). This article is intended as an overview of tensor networks for beginners looking to perform computations. We focus on the tools necessary for the concrete implementation of algorithms. Four basic operations (reshaping, permutation, contraction, and decomposition) that are commonly used in tensor network algorithms are covered. The article also briefly reviews diagrammatic notation, matrix product states (MPS), matrix product operators (MPO), projected entangled pair states (PEPS), multi-scale entanglement renormalization ansatz (MERA), time evolution block decimation (TEBD), and tensor renormalization group (TRG).

quant-ph↗