graphix 0.3.4

Optimize and simulate measurement-based quantum computation

Graphix is a measurement-based quantum computing (MBQC) software package, featuring

• the measurement calculus framework with integrated graphical rewrite rules for Pauli measurement preprocessing
• circuit-to-pattern transpiler, graph-based deterministic pattern generator and manual pattern generation
• flow, gflow and pauliflow finding tools and graph visualization based on flows (see below)
• statevector, density matrix and tensornetwork pattern simulation backends
• QPU interface and fusion network extraction tool
• new: efficient implementation of fast O(N^3) pauli-flow finding algorithm

Installation

Install graphix with pip:

pip install graphix

Install together with extra packages:

pip install graphix[extra]

Using graphix

generating pattern from a circuit

from graphix import Circuit

circuit = Circuit(4)
circuit.h(0)
...
pattern = circuit.transpile().pattern
pattern.standardize()
pattern.shift_signals()
pattern.draw_graph(flow_from_pattern=False)

See our example code to generate this pattern. Arrows indicate the causal flow of MBQC and dashed lines are the other edges of the graph. the vertical dashed partitions and the labels 'l:n' below indicate the execution layers or the order in the graph (measurements should happen from left to right, and nodes in the same layer can be measured simultaneously), based on the partial order associated with the (maximally-delayed) flow.

preprocessing Pauli measurements (Clifford gates)

pattern.perform_pauli_measurements()
pattern.draw_graph()

(here, the visualization is based on generalized flow).

simulating the pattern

state_out = pattern.simulate_pattern(backend="statevector")

and more..

• See demos showing other features of graphix.
• Read the tutorial for more usage guides.
• For theoretical background, read our quick introduction into MBQC and LC-MBQC.
• Full API docs is here.

Related packages

• graphix-stim-backend: stim backend for efficient Clifford pattern simulation
• graphix-symbolic: parameterized patterns with symbolic simulation
• graphix-ibmq: pattern transpiler for IBMQ / qiskit
• graphix-perceval: pattern transpiler for Quandela's perceval simulator and QPU
• graphix-qasm-parser: a plugin for parsing OpenQASM circuit.
• swiflow: rust-based implementation of flow-finding algorithms.
• graphqomb: modular graph state compiler for fault-tolerant MBQC and more.

Projects using graphix

• veriphix: verified blind quantum computation and benchmarking.
• optyx: ZX-based software for networked quantum computing.

Citing

Zenodo: https://doi.org/10.5281/zenodo.7861382

arXiv: https://doi.org/10.48550/arXiv.2212.11975

Contributing

We use GitHub issues for tracking feature requests and bug reports.

Discussion channels

• Our Slack channel, for regular discussions and questions: https://graphix-org.slack.com

• Please visit Unitary Foundation's Discord server, where you can find a channel for graphix.

Maintainers (alphabetical order)

• Masato Fukushima (University of Tokyo, Fixstars Amplify)
• Maxime Garnier (Inria Paris)
• Emlyn Graham (Inria Paris)
• Thierry Martinez (Inria Paris)
• Pranav Nair (Inria Paris)
• Sora Shiratani (University of Tokyo, Fixstars Amplify)
• Shinichi Sunami (University of Oxford)
• Mateo Uldemolins (Inria Paris)

Acknowledgements

Graphix was founded in 2022 by Shinichi Sunami (University of Oxford) and Masato Fukushima (University of Tokyo, Fixstars Amplify) with support from Fixstars Amplify and Unitary Foundation, and later joined by Daichi Sasaki, Yuki Watanabe and Sora Shiratani (University of Tokyo, Fixstars Amplify).

Since 2023, Graphix team is joined by Qode group of the QAT team, co-hosted by Inria and ENS, who develops and maintains the library.

   

Special thanks also to HQI.

License

Apache License 2.0