Interfaces¶

Interfaces to external packages all operate under the same overall design strategy. They are each subclassed off Interface and are each a singleton class with a standard set of functions to enable conversion to and from True-Q™. Each interface defines a default_config(), which is a Config object defining gates which True-Q™ can convert into.

However, these default config objects typically have multiple GateFactorys that implement the same Gate for different parameters. For example Identity is a special case of both Z(phi) and X(phi) when phi=0. Each interface fixes this degeneracy by preferentially converting using the ordering of the factories present in a set config. What this means is that if the config for a given interface contains both a GateFactory for an X(phi) rotation followed by a factory for an I gate, and a True-Q™ gate representing the identity is converted using that interface, then the interface will preferentially convert the gate into a X(0) gate in the external format.

The currently set config may be accessed on any interface via the get_config() method, which will return a config containing the factories in the preferred order of conversion. This config may be set through the set_config().

When an interface is successfully loaded, Circuit gets additional functionality in the form of a Circuit.to_<interface>(), which will convert the True-Q™ circuit representation into the representation of the external package. By altering the set config for the specified package, the user can select specific gates to build into by default.

class trueq.interface.metadata.Metadata(mapping=None)¶

Stores data not recorded in Circuit objects to enable lossless conversion.

Parameters: mapping (dict) – A dictionary mapping True-Q™ qubit labels to the indexing of the target framework.

Interface¶

class trueq.interface.base.Interface¶

Base class for interfaces to external software libraries or standards.

Each interface maintains a Config containing a collection of GateFactorys that the interface may build into. A complete collection of all gate conversions supported may be found in the default_config() for the respective interface.

In addition to this default config, the interfaces each keep track of a subset of this config which defines the current gates they will preferentially build into. This active subset of the default config may be manipulated via the get_config() and set_config() defined by each interface.

When there is a possible degeneracy in the conversions, such as when converting an identity gate into either an I gate or X(0) in the external standard, the first usable factory listed in the currently set config is used.

The interfaces rely on introspection of the external software to build the possible gate conversions, in many cases when new gates are added to an external package the True-Q™ interface will automatically support the conversion without any required changes.

All software interfaces to True-Q™ are subclassed off of this class.

abstract classmethod to_trueq_circ(other_circuit)¶

Converts a circuit defined through the external standard into a Circuit.

Parameters: other_circuit – A circuit which is defined in the external standard.
Return type: Circuit

abstract classmethod pair_from_trueq_circ(circuit, passes=None)¶

Returns a pair (compiled_circuit, external_format) where compiled_circuit has type Circuit and is the given circuit after it has been compiled by this interface according to get_config() and the given passes, but immediately before it has been converted into external_format which has circuit type of the external library.

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

tuple

classmethod from_trueq_circ(circuit, passes=None)¶

Converts a Circuit into the external standard, equal to the second element of a call to pair_from_trueq_circ().

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

abstract classmethod to_trueq_gate(other_gate)¶

Converts a gate defined through the external standard into a NativeGate.

Parameters: other_gate – A gate which is defined in the external standard.
Return type: NativeGate

abstract classmethod from_trueq_gate(labels, gate)¶

Converts a Gate into an external standard.

Parameters

labels (tuple) – A tuple of labels on which the gate operates.
gate (Gate) – The Gate to be converted.

abstract classmethod set_config(config)¶

Sets the config containing the gates which the interface should build into, this config must be a subset of the default_config().

Parameters: config (Config) – The Config which defines the gates to build into.

classmethod get_config()¶

Gets the current configuration which defines the gates which the interface builds into.

Return type: Config

abstract classmethod default_config()¶

Returns a config which contains all of the available gate conversions present in the interface.

Gates present in this config are defined exactly one-to-one equivalent to the gate in the external standard, including the naming convention for the gate and parameters (if present).

Return type: Config

Cirq¶

class trueq.interface.cirq¶

classmethod set_config(config=None)¶

Defines the gates which the Cirq interface should build into, accepting any of the following

A subset of the config defined in the default_config().
None, which will set the config to the default_config().
A list of factory names which are a subset of the factory names in the
default_config().

import trueq as tq
# convert strictly into these gates:
tq.interface.cirq.set_config(["CX", "XPowGate", "ZPowGate"])

# reset and allow any gates available
tq.interface.cirq.set_config()

Parameters: config (Config | NoneType | list) – The Config which defines the gates to build into.

classmethod to_trueq_circ(cirq_circ)¶

Converts a cirq.Circuit into a Circuit.

This returns a tuple of the True-Q™ circuit and a CirqMetadata which contains information present in the Cirq circuit which is not represented in True-Q™ circuits and is required for the reverse conversion (see from_trueq_circ()).

Note

Measurements are always placed at the end of the circuit.

Currently only supports Cirq circuits using GridQubits.

Parameters: cirq_circ – A Cirq representation of a circuit.
Type: cirq.Circuit
Return type: tuple of (Circuit, CirqMetadata)

classmethod pair_from_trueq_circ(circuit, metadata=None, passes=None, device=None, join_meas=True)¶

Returns a pair (compiled_circuit, cirq_circuit) where compiled_circuit has type Circuit and is the given circuit after it has been compiled by this interface according to get_config() and the given passes, but immediately before it has been converted into cirq_circuit which has type cirq.Circuit.

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

tuple

classmethod from_trueq_circ(circuit, metadata=None, passes=None, device=None, join_meas=True)¶

Converts a Circuit into a cirq.Circuit.

import trueq as tq

# A round trip, to cirq and back
circ = tq.Circuit([{(0, 1): tq.Gate.cx}])
cirq_circ = tq.interface.cirq.from_trueq_circ(circ)
cirq_circ.to_trueq().draw()

Note

By default, arbitrary single-qubit operations are decomposed using the ZXZ QubitMode by default. This may be changed by setting the config to have another mode.

If a gate is provided which is not directly equivalent to a single gate in get_config(), it is decomposed using the Compiler.

PyQuil¶

class trueq.interface.pyquil¶

classmethod set_config(config=None)¶

Defines the gates which the PyQuil interface should build into, accepting any of the following

A subset of the config defined in the default_config().
None, which will set the config to the default_config().
A list of factory names which are a subset of the factory names in the
default_config().

import trueq as tq
# convert strictly into these gates:
tq.interface.pyquil.set_config(["CNOT", "RX", "RZ"])

# reset and allow any gates available
tq.interface.pyquil.set_config()

Parameters: config (Config) – The Config which defines the gates to build into.

classmethod to_trueq_circ(program)¶

Converts a pyquil.Program into a Circuit.

This returns a tuple of the True-Q™ circuit and a PyQuilMetadata which contains information present in the PyQuil program which is not represented in True-Q™ circuits and is required for the reverse conversion (see from_trueq_circ()).

Parameters: program (pyquil.Program) – The pyquil program to be converted into a True-Q™ circuit.
Return type: tuple of (Circuit, PyQuilMetadata)

classmethod pair_from_trueq_circ(circuit, metadata=None, passes=None)¶

Returns a pair (compiled_circuit, pyquil_circuit) where compiled_circuit has type Circuit and is the given circuit after it has been compiled by this interface according to get_config() and the given passes, but immediately before it has been converted into pyquil_circuit which has type pyquil.Program.

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

tuple

classmethod from_trueq_circ(circuit, metadata=None, passes=None)¶

Converts a Circuit into a pyquil.Program.

import pyquil
import trueq as tq

circ = tq.Circuit([{(0, 1): tq.Gate.cx}])

ext_circ = tq.interface.pyquil.from_trueq_circ(circ)

ext_circ.to_trueq().draw()

Qiskit¶

class trueq.interface.qiskit¶

classmethod set_config(config=None)¶

Defines the gates which the Qiskit interface should build into, accepting any of the following:

A subset of the config defined in the default_config().
None, which will set the config to the default_config().
A list of factory names which are a subset of the factory names in the
default_config().

import trueq as tq
# convert strictly into these gates:
tq.interface.qiskit.set_config(["CXGate", "RXGate", "RZGate"])

# reset and allow any gates available
tq.interface.qiskit.set_config()

Parameters: config (Config) – The Config which defines the gates to build into.

classmethod to_trueq_circ(other_circuit)¶

Converts a qiskit.QuantumCircuit into a Circuit.

This returns a tuple of the True-Q™ circuit and a QiskitMetadata which contains information present in the Qiskit circuit which is not represented in True-Q™ circuits and is required for the reverse conversion (see from_trueq_circ()).

Note

Barriers are not preserved, and measurements are always placed at the end of the circuits.

Parameters: other_circuit (qiskit.QuantumCircuit) – A Qiskit circuit representation of a circuit.
Return type: tuple of (Circuit, QiskitMetadata)

classmethod pair_from_trueq_circ(circuit, metadata=None, passes=None)¶

Returns a pair (compiled_circuit, qiskit_circuit) where compiled_circuit has type Circuit and is the given circuit after it has been compiled by this interface according to get_config() and the given passes, but immediately before it has been converted into qiskit_circuit which has type qiskit.circuit.quantumcircuit.QuantumCircuit.

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

tuple

classmethod from_trueq_circ(circuit, metadata=None, passes=None)¶

Converts a Circuit into an qiskit.circuit.quantumcircuit.QuantumCircuit.

See also pair_from_trueq_circ.

Parameters

circuit (Circuit) – The Circuit to be converted.
metadata (QiskitMetadata) – Metadata required to accurately reproduce the original Qiskit circuit.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

qiskit.circuit.quantumcircuit.QuantumCircuit

classmethod to_trueq_gate(other_gate)¶

Converts a Qiskit gate into a NativeGate.

Parameters: other_gate – A gate which is defined by the external standard.
Returns: Operation

classmethod from_trueq_gate(labels, gate)¶

Converts a Gate into a Qiskit gate.

Parameters

labels (tuple) – A tuple of labels on which the gate operates.
gate (Gate) – The Gate to be converted.

classmethod default_config()¶

Returns a Config containing a definition of all Qiskit gates which True-Q™ is able to convert into.

Returns: Config

classmethod qiskit_dag_to_tq_circ(dag)¶

Takes a Qiskit DAG and convert it into a True-Q™ Circuit.

Note

Barriers are not preserved, and measurements are always placed at the end of the circuits.

Parameters: dag (qiskit.dagcircuit.DAGCircuit) – A Qiskit DAG representation of a circuit.
Returns: the exact input required by from_trueq_circ(), see that function for a full description of values returned.
Return type: tuple of (Circuit, QiskitMetadata)

classmethod set_from_backend(backend)¶

Sets the active config for the interface to one compatible with the target backend.

This returns the active config after it is set.

import qiskit as qk
from trueq.interface import qiskit as tqk
from qiskit.test import ibmq_mock

tqk.set_from_backend(ibmq_mock.backend_mocks.FakeMelbourne())

Mode: ZXZXZ
Dimension: 2
Gates:
- U2Gate(phi, lam):
    Hamiltonian:
    - [Z, 57.29577951308235*phi]
    - [X, 57.29577951308229*lam]
    - [Y, 90.0]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- CXGate:
    Matrix:
    - [1.0, 0.0, 0.0, 0.0]
    - [0.0, 1.0, 0.0, 0.0]
    - [0.0, 0.0, 0.0, 1.0]
    - [0.0, 0.0, 1.0, 0.0]
    Involving: {'(1, 0)': (), '(1, 2)': (), '(2, 3)': (), '(4, 3)': (), '(4, 10)': (),
      '(5, 4)': (), '(5, 6)': (), '(5, 9)': (), '(6, 8)': (), '(7, 8)': (), '(9, 8)': (),
      '(9, 10)': (), '(11, 3)': (), '(11, 10)': (), '(11, 12)': (), '(12, 2)': (),
      '(13, 1)': (), '(13, 12)': ()}
- U3Gate(theta, phi, lam):
    Hamiltonian:
    - [Z, 57.295779513082316*phi]
    - [Y, 57.295779513082294*theta]
    - [Z, 57.295779513082735*lam]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- IGate:
    Hamiltonian:
    - [I, 0]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- U1Gate(theta):
    Hamiltonian:
    - [Z, 57.29577951308232*theta]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}

Parameters: backend (qiskit.provider.BaseBackend) – A backend from Qiskit.
Return type: Config

classmethod config_from_backend(backend)¶

Takes an IBM Backend and converts it into a Config object.

The resulting config will contain the chip topology constraints as well as the gates themselves.

import qiskit as qk
from trueq.interface import qiskit as tqk
from qiskit.test import ibmq_mock

tqk.config_from_backend(ibmq_mock.backend_mocks.FakeMelbourne())

Mode: ZXZXZ
Dimension: 2
Gates:
- U2Gate(phi, lam):
    Hamiltonian:
    - [Z, 57.29577951308235*phi]
    - [X, 57.29577951308229*lam]
    - [Y, 90.0]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- CXGate:
    Matrix:
    - [1.0, 0.0, 0.0, 0.0]
    - [0.0, 1.0, 0.0, 0.0]
    - [0.0, 0.0, 0.0, 1.0]
    - [0.0, 0.0, 1.0, 0.0]
    Involving: {'(1, 0)': (), '(1, 2)': (), '(2, 3)': (), '(4, 3)': (), '(4, 10)': (),
      '(5, 4)': (), '(5, 6)': (), '(5, 9)': (), '(6, 8)': (), '(7, 8)': (), '(9, 8)': (),
      '(9, 10)': (), '(11, 3)': (), '(11, 10)': (), '(11, 12)': (), '(12, 2)': (),
      '(13, 1)': (), '(13, 12)': ()}
- U3Gate(theta, phi, lam):
    Hamiltonian:
    - [Z, 57.295779513082316*phi]
    - [Y, 57.295779513082294*theta]
    - [Z, 57.295779513082735*lam]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- IGate:
    Hamiltonian:
    - [I, 0]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}
- U1Gate(theta):
    Hamiltonian:
    - [Z, 57.29577951308232*theta]
    Involving: {'(1,)': (), '(0,)': (), '(2,)': (), '(3,)': (), '(4,)': (), '(10,)': (),
      '(5,)': (), '(6,)': (), '(9,)': (), '(8,)': (), '(7,)': (), '(11,)': (), '(12,)': (),
      '(13,)': ()}

Parameters: backend (qiskit.provider.BaseBackend) – A backend from Qiskit.
Returns: A config which contains gates and topology of a given backend.
Return type: Config

class trueq.interface.Executor(circuits, backend, filename=None, n_shots=128, max_submissions=None, overwrite=False, store_compiled=True)¶

A class to asynchronously submit a CircuitCollection to a qiskit.providers.Backend.

Since many Qiskit backends have a limit to the number of circuits that can be submitted in a single job, batching of circuits must be performed. True-Q™ circuit collections have support for batching circuits like this, see batch().

The Executor accepts a list of batched circuits (or just a list of circuits and batches them automatically), converts all circuits in each batch into the Qiskit circuit format, and submits each batch as a job to run on the specified backend. The executor then keeps track of the status of each job asynchronously, updating the status of each job periodically. The results of the input circuits are updated with results as each job finishes.

Note

The executor uses rich display inside of Jupyter notebooks. Make sure you have ipywidgets installed, and that you have enabled the extension

pip install ipywidgets
jupyter nbextension enable --py widgetsnbextension

Since most devices have a limit as to the number of jobs that can be active on a backend at any given time, the executor will only submit up to a max_submissions number of batches at any one time. Once a batch has run on the device, and has been removed from the job queue, the executor will place another batch into the queue until there are a total of max_submissions submitted jobs at any one time.

Here is a worked example, where a simple circuit is submitted to the IBMQ simulator backend:

import trueq as tq
import qiskit as qk
from qiskit.test import ibmq_mock

# loading IBMQ account and selecting a backend
# qk.IBMQ.load_account()
# provider = qk.IBMQ.get_provider()
# backend = provider.get_backend('ibmq_qasm_simulator')
# using a mock device as an example
backend = ibmq_mock.backend_mocks.FakeMelbourne()

# Generating a simple circuit of only the H gate
circs = tq.Circuit([{0: tq.Gate.h}, {0: tq.Meas()}])

# The handling of the submission of the circuits
ex = tq.interface.Executor(circs, backend)

# This is a blocking operation, and will block until all data has been acquired
ex.block()
circs.results.plot()

True-Q formatting will not be loaded without trusting this notebook or rerunning the affected cells. Notebooks can be marked as trusted by clicking "File -> Trust Notebook".

# A more complicated example using advanced batching options
circs2 = tq.make_srb((0, 1), [4, 64, 128])

# Batch the circuits into groups of 75, including readout calibration circuits
# in each batch
ro_circuits = tq.make_rcal([0, 1])
batches = circs2.batch(75, extra_circuits=ro_circuits)

ex2 = tq.interface.Executor(batches, backend)
ex2.block()
circs2.fit()

True-Q formatting will not be loaded without trusting this notebook or rerunning the affected cells. Notebooks can be marked as trusted by clicking "File -> Trust Notebook".

SRB Streamlined Randomized Benchmarking	Cliffords (0,) Key: backend: fake_melbourne labels: (0,) protocol: SRB twirl: Cliffords on [0, 1]	Cliffords (1,) Key: backend: fake_melbourne labels: (1,) protocol: SRB twirl: Cliffords on [0, 1]
${e}_{F}$ The probability of an error acting on the targeted systems during a random gate.	6.5e-06 (5.4e-05) 6.4928755194870735e-06, 5.425951858823649e-05	9.7e-05 (6.3e-05) 9.746327410081745e-05, 6.335613122765447e-05
${p}$ Decay parameter of the exponential decay $Ap^m$.	1.0e+00 (7.2e-05) 0.9999913428326407, 7.234602478431532e-05	1.0e+00 (8.4e-05) 0.9998700489678656, 8.447484163687263e-05
${A}$ SPAM parameter of the exponential decay $Ap^m$.	9.9e-01 (5.9e-03) 0.9927321418873655, 0.005908867450408166	1.0e+00 (7.3e-03) 1.0330498538503181, 0.007272679714811446
RCAL Readout Calibration	Confusion
(0,)	P(0 \| 0) = 1.000 P(1 \| 1) = 0.938 Confusion Matrix: Key: backend: fake_melbourne batch: 1 labels: (0, 1) protocol: RCAL
(1,)	P(0 \| 0) = 0.984 P(1 \| 1) = 0.945 Confusion Matrix: Key: backend: fake_melbourne batch: 1 labels: (0, 1) protocol: RCAL
RCAL Readout Calibration	Confusion
(0,)	P(0 \| 0) = 0.992 P(1 \| 1) = 0.961 Confusion Matrix: Key: backend: fake_melbourne batch: 0 labels: (0, 1) protocol: RCAL
(1,)	P(0 \| 0) = 0.969 P(1 \| 1) = 0.930 Confusion Matrix: Key: backend: fake_melbourne batch: 0 labels: (0, 1) protocol: RCAL

Parameters

circuits (CircuitCollection | Circuit | generator) –
The circuits which are to be submitted to the Qiskit backend. Depending on what is submitted here, there are several possible outcomes:
- Circuit
  It will be converted to a CircuitCollection. Results will be placed back inside of the original circuit object itself, and can be viewed as data is retrieved from the backend without having to wait.
- CircuitCollection
  It will be batched using the maximum allowed batch size for the given backend. For very long circuits this may result in errors due to limited memory of the classical hardware which control the backend. If errors occur during the submission of a CircuitCollection, it is recommended to resubmit using the third submission type. Results are placed back inside of the original circuits as data is retrieved from the server.
- batch
  The circuits will be batched exactly as given by the batch generator. This can be used to submit automatic readout correction circuits, see above example or batch() for more details about possible options. Results are placed inside original circuits as data comes in, note that some batching options may result in additional circuits being added to the collection.
backend (qiskit.providers.Backend) – The Qiskit backend to which the circuits will be submitted.
filename (str) – Optional filename where results will be saved. If this is provided the Executor will also be able to partially recover from a crash of the Python session, loss of internet, or other interruption in data acquisition.
n_shots (int) – The number of shots of the circuits to be aquired.
max_submissions (int | None) – The maximum number of batches allowed in a backend queue at any one time. If None, we will ask the backend how many remaining queue positions you are allowed, which will depend on whether you have other jobs running.
overwrite (bool) – Whether to overwrite the file on disk if it exists before these circuits are executed. Note that regardless of the value of this flag, after this first check, the file on disk will continually be updated as new results come in.
store_compiled (bool) – Whether to store (and save if filename is present) the compiled version of each circuit rather than the circuits as they were provided to this constructor. The compiled circuits are True-Q™ circuits after they have been compiled into the native gates of the given backend, but before they have been converted into Qiskit circuits. Note that during Qiskit conversion full system barriers are added between cycles with unequal markers.

property n_batches¶

The number of batches the circuits have been divided into.

int

property backend¶

Qiskit backend where the circuits will be submitted.

Type: qiskit.providers.Backend

property batches¶

A list of CircuitCollections, each of which is submitted separately to the backend.

Type: list

block()¶: Blocks the main thread until all circuits have completed acquisition on the backend.

property circuits¶

All submitted circuits.

Type: CircuitCollection

property status¶

A text representation for the current status of all jobs.

This is a string equivalent of the html output provided in jupyter.

Type: str

save()¶: Saves the current CircuitCollection regardless of status of acquisition.

cancel()¶

Cancels all currently submitted jobs if possible.

It is recommended that this be run any time a submission mistake is made, as circuits that have been submitted to the backend will stay in queue otherwise.

class trueq.interface.metadata.QiskitMetadata(mapping=None, meas_pairs=None)¶

Stores data not recorded in Circuit objects to enable lossless conversion.

Parameters

mapping (dict) – A dictionary mapping True-Q™ qubit labels to qiskit.Qubits.
meas_pairs (list) – A list of pairs, each with type (qiskit.Qubit, qiskit.Bit), whose order matches that of meas_locs. This stores which classical register each quantum measurement is to be stored into.

QASM¶

class trueq.interface.QASM¶

Interface class to convert Circuits into the QASM 2.0 language.

QASM 2.0 only defines two possible gate operations natively:

U
This is an arbitrary single qubit unitary with three free parameters.
CX
This is a fixed two qubit gate, more commonly called CNOT.

A config containing these two gates is accessable via default_config(). Any other gates used in a QASM file must be defined as a subroutine using these two fundamental gates.

The QASM interface supports Config objects which contain any one or two qubit fixed gates, as well as any of the following parameterized single qubit operations:

Pauli rotation on a single qubit.
u1_factory
u2_factory
u3_factory

This config the interface uses may be altered via set_config() and get_config(), and gates defined by it will automatically be converted to QASM formatted gate definitions inside of the QASM header.

A commonly available standard library of QASM gates called QELIB1 can be accessed through QELIB1().

import trueq as tq
circ = tq.make_xrb([0], [5], 1)[0]
print(circ.to_qasm())

// True-Q Version: 2.11.4
// Circuit Key:
// {'measurement_basis': 'X', 'n_random_cycles': 5, 'protocol': 'XRB', 'seq_label': 1144, 'twirl': ('trueq.Twirl', {'dim': 2, (0,): 'C'})}
OPENQASM 2.0;

qreg q[1];
creg c[1];
U(3.141592653589793,1.5707963267948966,0.0) q[0];
barrier q;
U(1.5707963267948966,-3.141592653589793,0.0) q[0];
barrier q;
U(3.141592653589793,0.0,0.0) q[0];
barrier q;
U(1.5707963267948966,-1.5707963267948966,0.0) q[0];
barrier q;
U(1.5707963267948966,-1.5707963267948966,0.0) q[0];
barrier q;
measure q[0] -> c[0];

Building into custom defined QASM gates:

import trueq as tq

zz_fact = tq.config.GateFactory.from_matrix("ZZ", tq.Gate.rp("ZZ", 90))
factories = [tq.interface.QASM.QELIB1().u1,
             tq.interface.QASM.QELIB1().u2,
             tq.interface.QASM.QELIB1().u3,
             zz_fact]
config = tq.Config(factories=factories)

tq.interface.QASM.set_config(config)

circ = tq.make_xrb([0], [5], 1)[0]
print(circ.to_qasm())

// True-Q Version: 2.11.4
// Circuit Key:
// {'measurement_basis': 'Y', 'n_random_cycles': 5, 'protocol': 'XRB', 'seq_label': 2610, 'twirl': ('trueq.Twirl', {'dim': 2, (0,): 'C'})}
OPENQASM 2.0;
gate u1(lambda) q { U(0,0,lambda) q; }
gate u2(phi,lambda) q { U(pi/2,phi,lambda) q; }
gate u3(theta,phi,lambda) q { U(theta,phi,lambda) q; }
gate ZZ q0, q1
{
    U(0.0,-3.141592653589793,0.0) q0;
    U(1.5707963267948966,-3.141592653589793,0.0) q1;
    CX q0,q1;
    U(0.0,1.5707963267948966,0.0) q0;
    U(1.5707963267948966,-1.5707963267948966,0.0) q1;
}
qreg q[1];
creg c[1];
u2(1.5707963267948966,0.0) q[0];
barrier q;
u2(1.5707963267948966,-1.5707963267948966) q[0];
barrier q;
u2(3.141592653589793,-1.5707963267948966) q[0];
barrier q;
u2(0.0,3.141592653589793) q[0];
barrier q;
u3(3.141592653589793,0.0,0.0) q[0];
barrier q;
measure q[0] -> c[0];

classmethod set_config(config=None)¶

Sets the Config which contains any one or two qubit fixed gates, as well as any of the following parameterized single qubit operations:

Pauli rotation on a single qubit.

u1_factory

u2_factory

u3_factory

The provided config will automatically be converted to QASM compatible gate definitions inside of the QASM header and may be used inside of the QASM file itself.

Parameters: config (Config) – The config containing gate definitions that are desired for the QASM representation.

classmethod get_header()¶

Returns the lines of the header currently associated with the currently desired QASM representation as defined by the get_config().

Return type: str

classmethod from_trueq_gate(labels, gate)¶

Converts a Gate into a QASM gate in the currently specified QASM respresentation as defined by get_config().

Parameters

labels (tuple) – A tuple of labels on which the gate operates.
gate (Gate) – The Gate to be converted.

Return type

str

Raises

ValueError – If conversion is not possible.

classmethod pair_from_trueq_circ(circuit, custom_header=None, include_auto_header=True, passes=None)¶

Returns a pair (compiled_circuit, qasm_str) where compiled_circuit has type Circuit and is the given circuit after it has been compiled by this interface according to get_config() and the given passes, but immediately before it has been converted into qasm_str which is a QASM 2.0 string.

Parameters

circuit (Circuit) – The Circuit to be converted.
passes (list) – A list of compiler passes to use during the conversion, this defaults to the HARDWARE_PASSES.

Return type

tuple

classmethod from_trueq_circ(circuit, custom_header=None, include_auto_header=True, passes=None)¶

Converts a Circuit into a QASM 2.0 text string.

Circuits converted to QASM will use the gate definitions found in the get_config(). These definitions will by default be included at the top of the output QASM file (this behavior can be disabled by setting include_auto_header=False). Additional lines in the header can be included by providing a list of strings to custom_header, which will be appended as new lines at the top of the file.