This package provides function optimize_kahypar for finding tensor network contraction orders. So far, it is only tested on Ubuntu. Mac OS and Windows users have to wait for the Binary issue of KaHyPar being solved.
OMEinsumContractionOrders is a Julia Language package. To install OMEinsumContractionOrders, please open Julia's interactive session (known as REPL) and press ] key in the REPL to use the package mode, then type the following command
For stable release
pkg> add https://github.com/kahypar/KaHyPar.jl.git#master
pkg> add OMEinsumContractionOrdersContract a tensor network
julia> using OMEinsum, OMEinsumContractionOrders, LightGraphs
julia> function random_regular_eincode(n, k; optimize=nothing)
g = LightGraphs.random_regular_graph(n, k)
ixs = [minmax(e.src,e.dst) for e in LightGraphs.edges(g)]
return EinCode((ixs..., [(i,) for i in LightGraphs.vertices(g)]...), ())
end
julia> code = random_regular_eincode(200, 3);
julia> optcode = optimize_kahypar(code, uniformsize(code, 2); sc_target=30, max_group_size=50);
julia> OMEinsum.timespace_complexity(code, uniformsize(code, 2))
(200.0, 0.0)
julia> OMEinsum.timespace_complexity(optcode, uniformsize(code, 2))
(38.0290167456887, 26.0)If you find this package useful in your research, please cite the following papers
@misc{Pan2021,
title={Simulating the Sycamore quantum supremacy circuits},
author={Feng Pan and Pan Zhang},
year={2021},
eprint={2103.03074},
archivePrefix={arXiv},
primaryClass={quant-ph}
}
@article{Gray2021,
title={Hyper-optimized tensor network contraction},
volume={5},
ISSN={2521-327X},
url={http://dx.doi.org/10.22331/q-2021-03-15-410},
DOI={10.22331/q-2021-03-15-410},
journal={Quantum},
publisher={Verein zur Forderung des Open Access Publizierens in den Quantenwissenschaften},
author={Gray, Johnnie and Kourtis, Stefanos},
year={2021},
month={Mar},
pages={410}
}
Jin-Guo Liu and Pan Zhang