The Quaspy library for Python

The Quaspy (Quantum algorithm simulations in Python) library for Python contains modules that implement:

• Simulators for the quantum part of Shor's order-finding algorithm [Shor94], modified as in [E24], and the classical post-processing from [E24] that recovers the order in a single run with very high success probability.

  Furthermore, these modules implement simulators for the quantum part of Seifert's variation [Seifert01] of Shor's order-finding algorithm [Shor94], as described as in [E24t] and [E21], and with the classical post-processing from [E24t] and [E21] with supporting functions from [E24] that efficiently recovers the order, or a positive integer multiple of the order, in multiple runs when making tradeoffs.

  Examples | Documentation |

• Simulators for factoring general integers via order-finding, and the classical post-processing from [E21b] and [E24] that factors any integer completely in a single order-finding run with very high success probability.

  Examples | Documentation |

• Simulators for the quantum part of Shor's algorithm for computing general discrete logarithms [Shor94], modified as in [E19p], and the classical post-processing from [E19p] that recovers the logarithm given the order in a single run with very high probability of success, and that also efficiently recovers the logarithm given the order when making tradeoffs.

  Examples | Documentation |

• Simulators for the quantum part of Ekerå–Håstad's algorithm for computing short discrete logarithms [EH17], modified as in [E20] and [E23p], and the classical post-processing from [E23p] that recovers the logarithm in a single run with very high probability of success. This algorithm does not require the order to be known.

  Furthermore, these modules implement the classical post-processing from [E20] that efficiently recovers the logarithm in multiple runs of the quantum part of Ekerå–Håstad's algorithm when making tradeoffs.

  Examples | Documentation |

• Simulators for factoring RSA integers via short discrete logarithms, by using the reduction in [EH17], modified as in [E20] and [E23p], and the classical post-processing from [E23p] that factors random RSA integers in a single run of the quantum part of Ekerå–Håstad's algorithm with very high probability of success.

  Furthermore, these modules implement the classical post-processing from [E20] that efficiently factors random RSA integers in multiple runs of the quantum part of Ekerå–Håstad's algorithm when making tradeoffs.

  Examples | Documentation |

All modules, classes, methods and functions in Quaspy are documented using Python docstrings.

Note that Quaspy implements basic support for tradeoffs via a native Python implementation of LLL that is stable and resasonable performant. See also the Qunundrum repository with its suite of MPI programs that implements support for tradeoffs via LLL and BKZ as implemented by fpLLL. Note furthermore that portions of Quaspy are inherited from the Factoritall repository.

Quaspy is a work in progress, and may be subject to major changes without prior notice. Quaspy was developed for academic research purposes. It grew out of our research project in an organic manner as research questions were posed and answered. It is distributed "as is" without warranty of any kind, either expressed or implied. For further details, see the license.

Notes on known memory leaks in gmpy2

Please note that Quaspy depends on GMP and MPFR via the gmpy2 wrapper for Python3, and that there is a memory leak issue in gmpy2 that currently impacts Quaspy. We have reported this issue, as have others before us, and there is a fairly trivial fix for it, but the maintainers of gmpy2 have thus far delayed the publication of said fix to Pip3 for reasons unknown to us.

Prerequisites

To install Python under Ubuntu 24.04 LTS, along with required dependencies, execute:

$ sudo apt install python3 python3-pip python3-venv
$ sudo apt install libgmp-dev libmpfr-dev libmpc-dev

For other Linux and Unix distributions, or operating systems, you may need to download Python and install it manually along with the required dependencies.

Installing the library

To install the latest pre-release of Quaspy via Pip3, execute:

$ pip3 install --pre quaspy

You may also install Quaspy directly from this repository, by executing:

$ pip3 install dist/quaspy-1.0.0a0-py3-none-any.whl

If you get get an error message to the effect that the environment is externally managed, you first need to setup a virtual environment, by executing:

$ python3 -m venv quaspy

You may then use quaspy/bin/pip3 and quaspy/bin/python3 to install and use Quaspy.

Examples

For examples that illustrate how to use Quaspy, please see the examples directory.

See also the documentation for Quaspy for help on how to use the library.

About and acknowledgments

The Quaspy library was developed by Martin Ekerå, in part at KTH, the Royal Institute of Technology, in Stockholm, Sweden. Valuable comments and advice were provided by Johan Håstad throughout the development process.

Funding and support was provided by the Swedish NCSA that is a part of the Swedish Armed Forces.