Source code for quantify_scheduler.operations.gate_library

# Repository: https://gitlab.com/quantify-os/quantify-scheduler
# Licensed according to the LICENCE file on the main branch
"""Standard gateset for use with the quantify_scheduler."""
from __future__ import annotations
from typing import Literal, Optional, Tuple, Hashable

import numpy as np

from .operation import Operation
from ..enums import BinMode


[docs] class Rxy(Operation): r""" A single qubit rotation around an axis in the equator of the Bloch sphere. This operation can be represented by the following unitary as defined in https://doi.org/10.1109/TQE.2020.2965810: .. math:: \mathsf {R}_{xy} \left(\theta, \varphi\right) = \begin{bmatrix} \textrm {cos}(\theta /2) & -ie^{-i\varphi }\textrm {sin}(\theta /2) \\ -ie^{i\varphi }\textrm {sin}(\theta /2) & \textrm {cos}(\theta /2) \end{bmatrix} Parameters ---------- theta Rotation angle in degrees, will be casted to the [-180, 180) domain. phi Phase of the rotation axis, will be casted to the [0, 360) domain. qubit The target qubit. """ def __init__( self, theta: float, phi: float, qubit: str, ): if not isinstance(theta, float): theta = float(theta) if not isinstance(phi, float): phi = float(phi) # this solves an issue where different rotations with the same rotation angle # modulo a full period are treated as distinct operations in the OperationDict # Here we map [0,360[ onto ]-180,180] so that X180 has positive amplitude theta = round(_modulo_360_with_mapping(theta), 8) phi = round(phi % 360, 8) tex = r"$R_{xy}^{" + f"{theta:.0f}, {phi:.0f}" + r"}$" plot_func = ( "quantify_scheduler.schedules._visualization.circuit_diagram.gate_box" ) theta_r = np.deg2rad(theta) phi_r = np.deg2rad(phi) # not all operations have a valid unitary description # (e.g., measure and init) unitary = np.array( [ [ np.cos(theta_r / 2), -1j * np.exp(-1j * phi_r) * np.sin(theta_r / 2), ], [ -1j * np.exp(1j * phi_r) * np.sin(theta_r / 2), np.cos(theta_r / 2), ], ] ) super().__init__(f"Rxy({theta:.8g}, {phi:.8g}, '{qubit}')") self.data["gate_info"] = { "unitary": unitary, "tex": tex, "plot_func": plot_func, "qubits": [qubit], "operation_type": "Rxy", "theta": theta, "phi": phi, } self._update() def __str__(self) -> str: gate_info = self.data["gate_info"] theta = gate_info["theta"] phi = gate_info["phi"] qubit = gate_info["qubits"][0] return f"{self.__class__.__name__}({theta=:.8g}, {phi=:.8g}, qubit='{qubit}')"
[docs] class X(Rxy): r""" A single qubit rotation of 180 degrees around the X-axis. This operation can be represented by the following unitary: .. math:: X180 = R_{X180} = \begin{bmatrix} 0 & -i \\ -i & 0 \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=180.0, phi=0, qubit=qubit) self.data["name"] = f"X {qubit}" self.data["gate_info"]["tex"] = r"$X_{\pi}$" self._update() def __str__(self) -> str: qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class X90(Rxy): r""" A single qubit rotation of 90 degrees around the X-axis. It is identical to the Rxy gate with theta=90 and phi=0 Defined by the unitary: .. math:: X90 = R_{X90} = \frac{1}{\sqrt{2}}\begin{bmatrix} 1 & -i \\ -i & 1 \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=90.0, phi=0.0, qubit=qubit) self.qubit = qubit self.data["name"] = f"X_90 {qubit}" self.data["gate_info"]["tex"] = r"$X_{\pi/2}$" self._update() def __str__(self) -> str: qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class Y(Rxy): r""" A single qubit rotation of 180 degrees around the Y-axis. It is identical to the Rxy gate with theta=180 and phi=90 Defined by the unitary: .. math:: Y180 = R_{Y180} = \begin{bmatrix} 0 & -1 \\ 1 & 0 \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=180.0, phi=90.0, qubit=qubit) self.data["name"] = f"Y {qubit}" self.data["gate_info"]["tex"] = r"$Y_{\pi}$" self._update() def __str__(self) -> str: qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class Y90(Rxy): r""" A single qubit rotation of 90 degrees around the Y-axis. It is identical to the Rxy gate with theta=90 and phi=90 Defined by the unitary: .. math:: Y90 = R_{Y90} = \frac{1}{\sqrt{2}}\begin{bmatrix} 1 & -1 \\ 1 & 1 \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=90.0, phi=90.0, qubit=qubit) self.data["name"] = f"Y_90 {qubit}" self.data["gate_info"]["tex"] = r"$Y_{\pi/2}$" self._update() def __str__(self) -> str: """ Returns a unique, evaluable string for unchanged data. Returns a concise string representation which can be evaluated into a new instance using :code:`eval(str(operation))` only when the data dictionary has not been modified. This representation is guaranteed to be unique. """ qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class Rz(Operation): r""" A single qubit rotation about the Z-axis of the Bloch sphere. This operation can be represented by the following unitary as defined in https://www.quantum-inspire.com/kbase/rz-gate/: .. math:: \mathsf {R}_{z} \left(\theta\right) = \begin{bmatrix} e^{-i\theta/2} & 0 \\ 0 & e^{i\theta/2} \end{bmatrix} Parameters ---------- theta Rotation angle in degrees, will be cast to the [-180, 180) domain. qubit The target qubit. """ def __init__(self, theta: float, qubit: str): if not isinstance(theta, float): theta = float(theta) # this solves an issue where different rotations with the same rotation angle # modulo a full period are treated as distinct operations in the OperationDict # Here we map [0,360[ onto ]-180,180] so that X180 has positive amplitude theta = _modulo_360_with_mapping(theta) tex = r"$R_{z}^{" + f"{theta:.0f}" + r"}$" plot_func = ( "quantify_scheduler.schedules._visualization.circuit_diagram.gate_box" ) theta_r = np.deg2rad(theta) # not all operations have a valid unitary description # (e.g., measure and init) unitary = np.array( [ [np.exp(-1j * theta_r / 2), 0], [0, np.exp(1j * theta_r / 2)], ] ) super().__init__(f"Rz({theta:.8g}, '{qubit}')") self.data["gate_info"] = { "unitary": unitary, "tex": tex, "plot_func": plot_func, "qubits": [qubit], "operation_type": "Rz", "theta": theta, } self._update() def __str__(self) -> str: gate_info = self.data["gate_info"] theta = gate_info["theta"] qubit = gate_info["qubits"][0] return f"{self.__class__.__name__}({theta=:.8g}, qubit='{qubit}')"
[docs] class Z(Rz): r""" A single qubit rotation of 180 degrees around the Z-axis. Note that the gate implements :math:`R_z(\pi) = -iZ`, adding a global phase of :math:`-\pi/2`. This operation can be represented by the following unitary: .. math:: Z180 = R_{Z180} = -iZ = e^{-\frac{\pi}{2}}Z = \begin{bmatrix} -i & 0 \\ 0 & i \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=180.0, qubit=qubit) self.data["name"] = f"Z {qubit}" self.data["gate_info"]["tex"] = r"$Z_{\pi}$" self._update() def __str__(self) -> str: qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class Z90(Rz): r""" A single qubit rotation of 90 degrees around the Z-axis. This operation can be represented by the following unitary: .. math:: Z90 = R_{Z90} = e^{-\frac{\pi/2}{2}}S = e^{-\frac{\pi/2}{2}}\sqrt{Z} = \frac{1}{\sqrt{2}}\begin{bmatrix} 1-i & 0 \\ 0 & 1+i \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, qubit: str): super().__init__(theta=90.0, qubit=qubit) self.data["name"] = f"Z_90 {qubit}" self.data["gate_info"]["tex"] = r"$Z_{\pi/2}$" self._update() def __str__(self) -> str: qubit = self.data["gate_info"]["qubits"][0] return f"{self.__class__.__name__}(qubit='{qubit}')"
[docs] class H(Operation): r""" A single qubit Hadamard gate. Note that the gate uses :math:`R_z(\pi) = -iZ`, adding a global phase of :math:`-\pi/2`. This operation can be represented by the following unitary: .. math:: H = Y90 \cdot Z = \frac{-i}{\sqrt{2}}\begin{bmatrix} 1 & 1 \\ 1 & -1 \\ \end{bmatrix} Parameters ---------- qubit The target qubit. """ def __init__(self, *qubits: str): tex = r"$H$" plot_func = ( "quantify_scheduler.schedules._visualization.circuit_diagram.gate_box" ) unitary = -1j / np.sqrt(2) * np.array([[1, 1], [1, -1]], dtype=complex) super().__init__(f"H, '{qubits}')") self.data["gate_info"] = { "unitary": unitary, "tex": tex, "plot_func": plot_func, "qubits": list(qubits), "operation_type": "H", } self._update() def __str__(self) -> str: qubits = map(lambda x: f"'{x}'", self.data["gate_info"]["qubits"]) return f'{self.__class__.__name__}({",".join(qubits)})'
[docs] class CNOT(Operation): r""" Conditional-NOT gate, a common entangling gate. Performs an X gate on the target qubit qT conditional on the state of the control qubit qC. This operation can be represented by the following unitary: .. math:: \mathrm{CNOT} = \begin{bmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 1 \\ 0 & 0 & 1 & 0 \\ \end{bmatrix} Parameters ---------- qC The control qubit. qT The target qubit """ def __init__(self, qC: str, qT: str): plot_func = "quantify_scheduler.schedules._visualization.circuit_diagram.cnot" super().__init__(f"CNOT ({qC}, {qT})") self.data.update( { "name": f"CNOT ({qC}, {qT})", "gate_info": { "unitary": np.array( [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 0, 1], [0, 0, 1, 0]] ), "tex": r"CNOT", "plot_func": plot_func, "qubits": [qC, qT], "symmetric": False, "operation_type": "CNOT", }, } ) self._update() def __str__(self) -> str: gate_info = self.data["gate_info"] qC = gate_info["qubits"][0] qT = gate_info["qubits"][1] return f"{self.__class__.__name__}(qC='{qC}',qT='{qT}')"
[docs] class CZ(Operation): r""" Conditional-phase gate, a common entangling gate. Performs a Z gate on the target qubit qT conditional on the state of the control qubit qC. This operation can be represented by the following unitary: .. math:: \mathrm{CZ} = \begin{bmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 1 & 0 \\ 0 & 0 & 0 & -1 \\ \end{bmatrix} Parameters ---------- qC The control qubit. qT The target qubit """ def __init__(self, qC: str, qT: str): plot_func = "quantify_scheduler.schedules._visualization.circuit_diagram.cz" super().__init__(f"CZ ({qC}, {qT})") self.data.update( { "name": f"CZ ({qC}, {qT})", "gate_info": { "unitary": np.array( [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, -1]] ), "tex": r"CZ", "plot_func": plot_func, "qubits": [qC, qT], "symmetric": True, "operation_type": "CZ", }, } ) self._update() def __str__(self) -> str: gate_info = self.data["gate_info"] qC = gate_info["qubits"][0] qT = gate_info["qubits"][1] return f"{self.__class__.__name__}(qC='{qC}',qT='{qT}')"
[docs] class Reset(Operation): r""" Reset a qubit to the :math:`|0\rangle` state. The Reset gate is an idle operation that is used to initialize one or more qubits. .. note:: Strictly speaking this is not a gate as it can not be described by a unitary. .. admonition:: Examples :class: tip The operation can be used in several ways: .. jupyter-execute:: from quantify_scheduler.operations.gate_library import Reset reset_1 = Reset("q0") reset_2 = Reset("q1", "q2") reset_3 = Reset(*[f"q{i}" for i in range(3, 6)]) Parameters ---------- qubits The qubit(s) to reset. NB one or more qubits can be specified, e.g., :code:`Reset("q0")`, :code:`Reset("q0", "q1", "q2")`, etc.. """ def __init__(self, *qubits: str): super().__init__(f"Reset {', '.join(qubits)}") plot_func = "quantify_scheduler.schedules._visualization.circuit_diagram.reset" self.data.update( { "name": f"Reset {', '.join(qubits)}", "gate_info": { "unitary": None, "tex": r"$|0\rangle$", "plot_func": plot_func, "qubits": list(qubits), "operation_type": "reset", }, } ) self._update() def __str__(self) -> str: qubits = map(lambda x: f"'{x}'", self.data["gate_info"]["qubits"]) return f'{self.__class__.__name__}({",".join(qubits)})'
[docs] class Measure(Operation): """ A projective measurement in the Z-basis. The measurement is compiled according to the type of acquisition specified in the device configuration. .. note:: Strictly speaking this is not a gate as it can not be described by a unitary. Parameters ---------- qubits The qubits you want to measure. acq_channel Only for special use cases. By default (if None): the acquisition channel specified in the device element is used. If set, this acquisition channel is used for this measurement. acq_index Index of the register where the measurement is stored. If None specified, this defaults to writing the result of all qubits to acq_index 0. By default None. acq_protocol : "SSBIntegrationComplex" | "Trace" | "TriggerCount" | \ "NumericalSeparatedWeightedIntegration" | \ "NumericalWeightedIntegration" | None, optional Acquisition protocols that are supported. If ``None`` is specified, the default protocol is chosen based on the device and backend configuration. By default None. bin_mode The binning mode that is to be used. If not None, it will overwrite the binning mode used for Measurements in the circuit-to-device compilation step. By default None. feedback_trigger_label : str The label corresponding to the feedback trigger, which is mapped by the compiler to a feedback trigger address on hardware, by default None. """ def __init__( self, *qubits: str, acq_channel: Hashable | None = None, acq_index: Tuple[int, ...] | int | None = None, # These are the currently supported acquisition protocols. acq_protocol: Optional[ Literal[ "SSBIntegrationComplex", "Trace", "TriggerCount", "NumericalSeparatedWeightedIntegration", "NumericalWeightedIntegration", "ThresholdedAcquisition", ] ] = None, bin_mode: BinMode | None = None, feedback_trigger_label: Optional[str] = None, ): # this if else statement a workaround to support multiplexed measurements (#262) # this snippet has some automatic behaviour that is error prone. # see #262 if len(qubits) == 1: if acq_index is None: acq_index = 0 else: if isinstance(acq_index, int): acq_index = [ acq_index, ] * len(qubits) elif acq_index is None: # defaults to writing the result of all qubits to acq_index 0. # note that this will result in averaging data together if multiple # measurements are present in the same schedule (#262) acq_index = list(0 for i in range(len(qubits))) plot_func = "quantify_scheduler.schedules._visualization.circuit_diagram.meter" super().__init__(f"Measure {', '.join(qubits)}") self.data.update( { "name": f"Measure {', '.join(qubits)}", "gate_info": { "unitary": None, "plot_func": plot_func, "tex": r"$\langle0|$", "qubits": list(qubits), "acq_channel_override": acq_channel, "acq_index": acq_index, "acq_protocol": acq_protocol, "bin_mode": bin_mode, "operation_type": "measure", "feedback_trigger_label": feedback_trigger_label, }, } ) self._update() def __str__(self) -> str: gate_info = self.data["gate_info"] qubits = map(lambda x: f"'{x}'", gate_info["qubits"]) acq_channel = gate_info["acq_channel_override"] acq_index = gate_info["acq_index"] acq_protocol = gate_info["acq_protocol"] bin_mode = gate_info["bin_mode"] feedback_trigger_label = gate_info["feedback_trigger_label"] return ( f'{self.__class__.__name__}({",".join(qubits)}, ' f"acq_channel={acq_channel}, " f"acq_index={acq_index}, " f'acq_protocol="{acq_protocol}", ' f"bin_mode={str(bin_mode)}, " f"feedback_trigger_label={feedback_trigger_label})" )
[docs] def _modulo_360_with_mapping(theta: float) -> float: """ Maps an input angle ``theta`` (in degrees) onto the range ``]-180, 180]``. By mapping the input angle to the range ``]-180, 180]`` (where -180 is excluded), it ensures that the output amplitude is always minimized on the hardware. This mapping should not have an effect on the qubit in general. -180 degrees is excluded to ensure positive amplitudes in the gates like X180 and Z180. Note that an input of -180 degrees is remapped to 180 degrees to maintain the positive amplitude constraint. Parameters ---------- theta : float The rotation angle in degrees. This angle will be mapped to the interval ``]-180, 180]``. Returns ------- float The mapped angle in degrees, which will be in the range ``]-180, 180]``. This mapping ensures the output amplitude is always minimized for transmon operations. Example ------- ``` >>> _modulo_360_with_mapping(360) 0.0 >>> _modulo_360_with_mapping(-180) 180.0 >>> _modulo_360_with_mapping(270) -90.0 ``` """ mapped_theta = -((-theta - 180) % 360) + 180 return mapped_theta