quantr

This crate is not production ready and so should not be considered stable, nor produce correct answers. It is still under heavy development and requires many more optimisations. Hence, it's likely that near future updates will induce breaking changes. Please always check answers with other simulations if you are intending to use quantr for projects.

A Rust library crate that builds, prints and simulates a quantum computer.

This crate allows the user to build a quantum circuit by adding columns of gates via various methods. Once the circuit has been built, then it can be simulated, which attaches the register |00..0> resulting in a superposition that can be measured.

For a brief example of using quantr, see the quick start guide which walks through an implementation of Grover's algorithm.

Defining features

• Aimed to be accessible for beginners in Rust.
• The distinction between physical observables and non-physical observables that can be extracted from the circuit is made clear, where the latter is still possible to retrieve.
• Prints the circuit diagram to the terminal, or saves it to a text file, as a UTF-8 string.
• Custom gates can be implemented easily by giving their explicit linear mappings on product states. This allows the user to avoid representing the gates as matrices.
• Attempts to minimise memory consumption by not using matrices nor sparse matrices, but instead uses functions to represent the linear mapping of gates. For the number of qubits in a circuit, n, the estimated memory that is required for quantr to simulate the circuit when n >= 16 is approximately the size of the state vector itself in Rust, 2**(n-6) KiB. For n < 16, the memory required is less than 1 MiB.
• Can simulate circuits up to ~18 qubits within a reasonable time.
• Only safe Rust code is used, and the only dependency is the fastrand crate and its sub-dependencies.

Usage

An example of simulating and printing a two qubit circuit:

use quantr::{Circuit, Gate, Printer, Measurement::Observable};

fn main() {

    let mut quantum_circuit: Circuit = Circuit::new(2).unwrap();
    
    quantum_circuit 
        .add_gates(&[Gate::H, Gate::Y]).unwrap()
        .add_gate(Gate::CNot(0), 1).unwrap();
    
    let mut printer = Printer::new(&quantum_circuit);
    printer.print_diagram();
    // The above prints the following:
    // ┏━━━┓     
    // ┨ H ┠──█──
    // ┗━━━┛  │  
    //        │  
    // ┏━━━┓┏━┷━┓
    // ┨ Y ┠┨ X ┠
    // ┗━━━┛┗━━━┛

    quantum_circuit.simulate();

    // Below prints the number of times that each state was observed 
    // over 500 measurements of superpositions.
    if let Ok(Observable(bin_count)) = quantum_circuit.repeat_measurement(500) {
        println!("[Observable] Bin count of observed states.");
        for (state, count) in bin_count {
            println!("|{}> observed {} times", state, count);
        }
    }

}

A more detailed example of using quantr is given in the quick start guide.

Limitations (currently)

• No noise consideration, or ability to introduce noise.
• No parallelisation option.
• No ability to add classical wires nor gates that measure a single wire of a quantum circuit.

Conventions

The ordering of the wires labelling the product states in the computational basis is defined as:

|a⟩ ──── 
|b⟩ ────  ⟺ |a,b,c,⋯⟩ ≡ |a⟩⊗|b⟩⊗|c⟩⊗⋯ 
|c⟩ ────
 ⋮    ⋮

When defining a custom function that depends on the position of control nodes to define gates (such as the CNot and Toffoli gates), it must be defined so that the most far right state of the product state, is assumed to be the gate that is 'activated'. In general, it is better to assume that the custom function doesn't define control nodes, but rather it extends the dimension of the function's domain. Lastly, the custom enum does not check if the mapping is unitary.

Documentation

The Quantr Book is planned to serve as extended documentation to quantr, such as explaining the motivations behind chosen algorithms. For now, it only contains the start guide.

For the online code documentation, please refer to crates.io. This can also be built and opened in your favourite web browser locally by cloning the project, moving into the directory, and running cargo doc --open.

Other quantum computer simulators

The website Are We Quantum Yet (checked 24/10/23) lists all things quantum computing in Rust.

A useful and very practical simulator for learning quantum computing is Quirk. It's a real-time online simulator that interfaces via drag-and-drop gates. Note that the labelling of the states in the computational basis in Quirk is reversed when compared to quantr's labelling of such states.

Licence

Quantr is licensed under the EUPL-1.2 or later. You may obtain a copy of the licence at https://joinup.ec.europa.eu/collection/eupl/eupl-text-eupl-12. A copy of the EUPL-1.2 licence in English is given in LICENCE.txt which is found in the root of this repository. Details of the licenses of third party software, and the quantr project, can be found in COPYRIGHT.txt.