qnn_circuit

class qnn_circuit(num_qubits=None, feature_map=None, ansatz=None)

Bases:

The qnn_circuit creates a QuantumCircuit that is a composition of a feature map and an ansatz circuit. Also returned are the feature map and ansatz parameters for use for inputs and weights as needed for a neural network, such as SamplerQNN.

If only the number of qubits is provided the real_amplitudes() ansatz and the zz_feature_map() feature map are used. If the number of qubits is 1 the z_feature_map() is used. If only a feature map is provided, the real_amplitudes() ansatz with the corresponding number of qubits is used. If only an ansatz is provided the zz_feature_map() with the corresponding number of qubits is used.

At least one parameter has to be provided. If a feature map and an ansatz is provided, the number of qubits must be the same.

Example:

from qiskit_machine_learning.circuit.library import qnn_circuit
qnn_qc, fm_params, anz_params = qnn_circuit(2)
qnn_qc.draw(fold=60)
#      ┌───┐┌─────────────┐     »
# q_0: ┤ H ├┤ P(2.0*x[0]) ├──■──»
#      ├───┤├─────────────┤┌─┴─┐»
# q_1: ┤ H ├┤ P(2.0*x[1]) ├┤ X ├»
#      └───┘└─────────────┘└───┘»
# «                                          ┌───┐»
# «q_0: ──────────────────────────────────■──┤ H ├»
# «     ┌──────────────────────────────┐┌─┴─┐├───┤»
# «q_1: ┤ P(2.0*(x[0] - π)*(x[1] - π)) ├┤ X ├┤ H ├»
# «     └──────────────────────────────┘└───┘└───┘»
# «     ┌─────────────┐                                     »
# «q_0: ┤ P(2.0*x[0]) ├──■──────────────────────────────────»
# «     ├─────────────┤┌─┴─┐┌──────────────────────────────┐»
# «q_1: ┤ P(2.0*x[1]) ├┤ X ├┤ P(2.0*(x[0] - π)*(x[1] - π)) ├»
# «     └─────────────┘└───┘└──────────────────────────────┘»
# «          ┌──────────┐     ┌──────────┐     ┌──────────┐»
# «q_0: ──■──┤ Ry(θ[0]) ├──■──┤ Ry(θ[2]) ├──■──┤ Ry(θ[4]) ├»
# «     ┌─┴─┐├──────────┤┌─┴─┐├──────────┤┌─┴─┐├──────────┤»
# «q_1: ┤ X ├┤ Ry(θ[1]) ├┤ X ├┤ Ry(θ[3]) ├┤ X ├┤ Ry(θ[5]) ├»
# «     └───┘└──────────┘└───┘└──────────┘└───┘└──────────┘»
# «          ┌──────────┐
# «q_0: ──■──┤ Ry(θ[6]) ├
# «     ┌─┴─┐├──────────┤
# «q_1: ┤ X ├┤ Ry(θ[7]) ├
# «     └───┘└──────────┘
print(fm_params)
# ParameterView([ParameterVectorElement(x[0]), ParameterVectorElement(x[1])])
print(anz_params)
# ParameterView([ParameterVectorElement(θ[0]), ParameterVectorElement(θ[1]),
#                ParameterVectorElement(θ[2]), ParameterVectorElement(θ[3]),
#                ParameterVectorElement(θ[4]), ParameterVectorElement(θ[5]),
#                ParameterVectorElement(θ[6]), ParameterVectorElement(θ[7])])

Although all arguments to qnn_circuit default to None at least one must be provided, to determine the number of qubits from.

If more than one is passed:

1) If num_qubits is provided the feature map and/or ansatz circuits supplied must be the same number of qubits.

2) If num_qubits is not provided the feature_map and ansatz must be set to the same number of qubits.

Parameters:

• num_qubits (int | None) – Number of qubits, a positive integer. Optional if feature_map or ansatz is provided, otherwise required. If not provided num_qubits defaults from the sizes of feature_map and/or ansatz.

• feature_map (QuantumCircuit | None) – A feature map. Optional if num_qubits or ansatz is provided, otherwise required. If not provided defaults to zz_feature_map() or z_feature_map() if num_qubits is determined to be 1.

• ansatz (QuantumCircuit | None) – An ansatz. Optional if num_qubits or feature_map is provided, otherwise required. If not provided defaults to real_amplitudes().

Returns:

The composed feature map and ansatz circuit, the feature map parameters and the ansatz parameters.

Raises:

QiskitMachineLearningError – If a valid number of qubits cannot be derived from the provided input arguments.