16 releases (5 breaking)
0.6.0 | Jul 21, 2024 |
---|---|
0.5.1 | Feb 4, 2024 |
0.4.0 | Dec 22, 2023 |
0.3.0 | Nov 17, 2023 |
#66 in Science
155KB
2.5K
SLoC
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. Please always check answers with other simulations if you are intending to use quantr for projects.
A Rust library crate that simulates gate-based quantum circuits with focus on memory efficiency and accessibility.
This crate allows the user to build quantum circuits 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. Quantr has primarily been built for the simulation of pure states.
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 mappings on product states. This allows the user to avoid representing the gates as matrices.
- Custom gates do not have to be unitary, allowing for some quantum channel to be implemented.
- Can simulate circuits up to ~16 qubits within a tractable time.
- Only safe Rust code is used, and the only dependencies are the fastrand (2.1.0) crate, num_complex (0.4.6), and their 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 ┠
// ┗━━━┛┗━━━┛
let simulated_circuit = quantum_circuit.simulate();
// Below prints the number of times that each state was observered
// over 500 measurements of superpositions.
if let Observable(bin_count) = simulated_circuit.measure_all(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, however this could be (albeit tediously) implemented through the custom gates.
- No parallelisation option.
- No ability to add classical wires nor gates that measure a single wire of a quantum circuit. Only one method is given that in effect attaches a measurement gate at the end of all qubit wires.
- Designed primarily for the simulation of pure state vectors.
Although, through the use of custom gates some quantum channels can be
implemented. For an example, see
examples/post_select.rs
.
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, it should be noted that there are no checks on the custom gate for being a valid quantum channel.
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, and some preliminary results of the memory efficiency of quantr.
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
Another, and more stable, quantum circuit simulator written in Rust is Spinoza.
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.
Dependencies
~295KB