States¶

Piquasso state classes represent the result of a simulation. In most workflows, you do not choose a state class directly. Instead, you choose a simulator, execute a program, and inspect the returned Result (e.g., result.state when no measurements are performed, otherwise result.samples or result.branches).

GaussianState

Returned by Gaussian simulations. It represents Gaussian states through phase-space data such as first and second moments.

Gaussian State

FockState and PureFockState

Returned by Fock-space simulations. They represent mixed or pure photonic states in a truncated Fock basis.

Fock States

PassiveState

Used by Boson Sampling simulations to represent the state before terminal particle number measurements.

Passive State

How states are obtained¶

The state type follows from the simulator used to execute the program.

GaussianSimulator

Returns a GaussianState when the program is executed without terminal sampling-only output.

PureFockSimulator

Returns a PureFockState represented by state-vector coefficients in the Fock basis.

FockSimulator

Returns a FockState represented by a density matrix in the Fock basis.

PassiveSimulator

Uses a PassiveState for Boson Sampling-style circuits and produces particle-number samples.

Basic usage¶

After executing a program without measurements, the final state can be accessed through result.state. For programs with measurements, inspect result.samples (finite shots) or result.branches (exact distribution with shots=None).

import numpy as np
import piquasso as pq

with pq.Program() as program:
   pq.Q() | pq.Vacuum()
   pq.Q(0) | pq.Displacement(r=0.2)
   pq.Q(0, 1) | pq.Beamsplitter(theta=np.pi / 4)

simulator = pq.GaussianSimulator(d=2)
result = simulator.execute(program)

state = result.state

print(type(state))
print(state.mean_photon_number())

Fock-state example¶

For non-Gaussian circuits, use a Fock-space simulator. The state is represented in a truncated Fock basis, controlled by Config(cutoff=...).

import piquasso as pq

with pq.Program() as program:
   pq.Q() | pq.Vacuum()
   pq.Q(0) | pq.Displacement(r=0.4)
   pq.Q(0) | pq.Kerr(xi=0.05)

simulator = pq.PureFockSimulator(
   d=1,
   config=pq.Config(cutoff=8),
)
result = simulator.execute(program)

state = result.state

print(type(state))
print(state.fock_probabilities)

Boson Sampling example¶

For Boson Sampling-style simulations, the main output is usually the collection of measurement samples.

import numpy as np
import piquasso as pq

interferometer = np.array(
   [
      [1 / np.sqrt(2), 1 / np.sqrt(2)],
      [1 / np.sqrt(2), -1 / np.sqrt(2)],
   ]
)

with pq.Program() as program:
   pq.Q() | pq.NumberState([1, 1])
   pq.Q() | pq.Interferometer(interferometer)
   pq.Q(all) | pq.ParticleNumberMeasurement()

simulator = pq.PassiveSimulator(d=2)
result = simulator.execute(program, shots=10)

print(result.samples)