ffsim 0.0.7

Faster simulations of fermionic quantum circuits.

ffsim

Faster simulations of fermionic quantum circuits.

What is ffsim?

ffsim is a software library for simulating fermionic quantum circuits that preserve particle number and the Z component of spin. This category includes many quantum circuits used for quantum chemistry simulations. By exploiting the symmetries and using specialized algorithms, ffsim can simulate these circuits much faster than a generic quantum circuit simulator.

Experimental disclaimer: ffsim is currently an experimental release. Breaking changes may be introduced without warning.

Features

• Fast simulation of fermionic quantum circuits that preserve particle number and the Z component of spin. ffsim supports the simulation of basic fermionic gates and includes specialized routines for simulation of molecular Hamiltonians in the "double-factorized" representation.
• Compatible with PySCF. State vectors use the same indexing convention as PySCF's fci module.

Limitations

• There is no support for operations that do not preserve particle number and the Z component of spin.

Supported platforms

ffsim is supported on Linux, macOS, and the Windows Subsystem for Linux (WSL). It is not supported on Windows.

Installation

From PyPI

ffsim is available on PyPI. It can be installed by running the command

pip install ffsim

From source

Installing ffsim from source requires the following system dependencies:

• A Rust compiler. See these instructions.

• A BLAS implementation. For example:

  • Arch Linux:

    sudo pacman -S blas-openblas
    

  • Fedora:

    sudo dnf install openblas-devel
    

  • Ubuntu:

    sudo apt install libopenblas-dev
    

Once these dependencies are satisfied, ffsim can be installed by running the command

pip install .

from the root directory of the code repository.

Development

To set up ffsim for development, install it from source in editable mode along with the development requirements:

pip install -e ".[dev]"

If you add or modify any Rust modules, rebuild them by running the command

maturin develop

If you are benchmarking the code, then pass the --release flag:

maturin develop --release

Tests and other code checks are managed using tox. To run the default tox environments, simply run

tox

To run a specific environment, for example, to run the lint checks, do

tox run -e lint

You can also use pytest to run the tests directly. For example,

pytest tests/