Getting started

What is Perceval?

Perceval is a toolbox containing generic functions and classes, built around an optimised native core (see exqalibur code reference).

It offers tools to:

Manipulate quantum states in the Fock space
- Pure states (FockState, StateVector)
- Mixed states (SVDistribution, DensityMatrix)
Build a linear optics Experiment containing
- A unitary Circuit composed of Unitary Components
- Some Non-unitary Components
- Feed-forward through Feed-forward Configurators
- Variable parameters and expressions to parametrise components
Display circuits and data (pdisplay), serialise them (serialization)
Define real-world noise parameters applied in the input, the linear-optics circuit and the photon detectors (Noise Model)
Simulate these experiments through different layers of simulations
- Perfect simulations with Simulation Back-ends
- Noisy and non-unitary simulations with the Simulator layer
Control the flow of quantum computations and choose where they are run:
- Locally with the SimulatedComputer, remotely with the RemoteComputer
- Manage your jobs with the ExecutionGroup

Installing Perceval

Perceval supports several Python versions (typically, those that are not in “end-of-life”). In a virtual environment of any Python supported version, a single pip command installs Perceval and all of its dependencies.

$ pip install perceval-quandela

Warning

Pay attention that the Python package name is “perceval-quandela” and not “perceval”

Once the above command succeeds, you can start typing code in your favorite IDE!

Hello world

The following example is a minimal code to simulate the Hong–Ou–Mandel effect on the user’s computer in a noisy situation, and retrieve both a sample count and exact probabilities computed by a strong simulation back-end.

>>> import perceval as pcvl
>>>
>>> experiment = pcvl.Experiment(pcvl.BS())          # In a perfect beam splitter experiment...
>>> experiment.with_input(pcvl.BasicState("|1,1>"))  # ... we inject one photon into each mode
>>> experiment.min_detected_photons_filter(1)  # Accept all output states containing at least 1 photon
>>>
>>> computer = pcvl.SimulatedComputer("SLOS")  # Use SLOS, a strong simulation back-end
>>> computer.noise = pcvl.NoiseModel(transmittance=0.05, indistinguishability=0.85)  # Define some noise level
>>>
>>> factory = pcvl.ExecutionFactory(computer, experiment)
>>> with computer.acquire():  # Good practice - may be used to warm the computer up
...     samples = factory.sample_count(10_000)['results']  # Ask to generate 10k samples, and get back only the raw results
...     probs = factory.probs()['results']  # Ask for the exact probabilities
>>> print(f"Samples: {samples}")
>>> print(f"Probabilities: {probs}")
Samples: {
  |2,0>: 117
  |0,2>: 147
  |1,0>: 4822
  |1,1>: 22
  |0,1>: 4892
}
Probabilities: {
  |2,0>: 0.011858974358974369
  |0,2>: 0.011858974358974369
  |1,1>: 0.0019230769230769245
  |1,0>: 0.48717948717948717
  |0,1>: 0.48717948717948717
}

Now that you can run some code, let’s continue with a tutorial to learn Perceval syntax.