Tensor Network Theory Library  Beta release 1.2.1
A library of routines for performing TNT-based operations
 All Data Structures Functions Variables Groups Pages
The TNT Library

This is the users' documentation for the Tensor Network Theory (TNT) library. The library contains routines for performing the tensor manipulations required by TNT-based algorithms, and for performing simulations using these algorithms.

The code is being developed in Dieter Jaksch's Quantum Systems Engineering Group at the Department of Physics, Oxford, with core software support from the Software Engineering Group at the STFC Rutherford Appleton Laboratory. The project has also received support from NAG (2012-2013) under the HECToR University dCSE scheme (see report).

This documentation contains reference for functions provided as part of the TNT library. For examples of how to use these functions, there is also example C code provided for download and in the last chapter of this documentation, which also explains how the code works.

Library Summary

The core library contains routines that are completely general i.e. they do not depend on the network or system geometry, or on the TNT algorithm being used. They can be used to build your own TNT algorithms. The functions in this library are available in the static library libtnt.a, and the source code is available on request.

In addition to the core library are additional libraries that do depend on the network geometry, and contain network functions that are the building blocks of specific TNT algorithms. These are available as static libraries and the source code is also available in the HTML version of the documentation. The only plug-in library currently available is libtntMps.a, which contains routines for manipulating matrix product states (MPS).

Citing the TNT library

If you use the library for work that results in a publication, please cite the library as Tensor Network Theory Library, Beta Version 1.2.0 (2016), S. Al-Assam, S. R. Clark, D. Jaksch and TNT Development team, www.tensornetworktheory.org and also cite our latest paper Tensor Network Theory - Part 1: Overview of core library and tensor operations, arXiv:1610.02244. The library currently requires you to link to a suitable linear algebra library. Currently supported options are

  • the NAG FORTRAN library (tested with fll6i24dcl).
  • A BLAS, LAPACK and ARPACK library.

The MATLAB library should also be linked to for input and output of variables, although more formats will be supported in later versions.

Contacting us

If you have any difficulties, please contact suppo.nosp@m.rt@t.nosp@m.ensor.nosp@m.netw.nosp@m.orkth.nosp@m.eory.nosp@m..org