Config¶
Contains the relevant configuration of a system, including which gates are available to be run, how many systems there are and the dimension of the individual systems. |
|
Constructs new |
|
A dictionary containing allowable sets of systems. |
|
Creates a Python function from a string containing numbers, variables, and standard math functions. |
Config¶
-
class
trueq.
Config
(mode='ZXZXZ', factories=None, dim=2)¶ Contains the relevant configuration of a system, including which gates are available to be run, how many systems there are and the dimension of the individual systems.
This class relies on
GateFactory
s to describe what gates are available on the hardware, see that documentation for more details.Here we use a convenience method to create a basic config which may be sufficient in many cases. This config will define a total of three possible gates on the system, two single qubit gates
Z
(a pauliZ
rotation) andX90
(a pauliX
rotation of 90 degrees), and a single two qubit gate defined by theentangler
keyword.import trueq as tq config = tq.Config.basic(entangler=tq.Gate.cx, mode="ZXZXZ") # we can access the GateFactory objects via the factories attribute. config.factories # for ease of use, individual factories are available as attributes of the config config.z # lets make an Z90 gate using this definition: config.z(phi=90) # The gate definitions can be accessed in several ways: assert config.z == config[0] == config['z'] == config.get("z")[0] assert config.cx == config[2] == config['cx']
These configs can be saved to a YAML file and loaded back in the future. This facilitates including topology and gate restrictions by specifying
Involving
keys in the YAML. Seefrom_yaml
andto_yaml
for more details.- Parameters
mode (
str
) – The single qubit decomposition mode, defaults to'ZXZXZ'
, seetrueq.math.decomposition.QubitMode
for more details.dim (
int
) – The dimension of computational system, defaults to 2 (qubits).factories (
Iterable
) – An iterable ofGateFactory
objects.
- Return type
- Raises
ValueError – If the provided
mode
is not defined inQubitMode
.
-
static
basic
(entangler=Gate.cx, mode='ZXZXZ', dim=2)¶ Creates a new
Config
object containing a fixed entangling gate and the Pauli rotation gates required to decompose single-qubit gates into the specified mode.import trueq as tq config = tq.Config.basic(entangler=tq.Gate.cnot, mode="ZXZXZ") # we can access the GateFactory objects via the factories attribute. config.factories # individual factories are available as attributes of the config config.z # lets make an Z90 gate using this definition: config.z(phi=90)
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".- Name:
-
- z(phi)
- Aliases:
-
- Gate.s
- Gate.sz
- Gate.cliff9
- Parameters:
-
- phi = 90
- Generators:
-
- 'Z': 90.0
- Matrix:
-
-
property
connectivity
¶ The connectivity graph of all multi-qubit gates in the config.
A
set
offrozenset
s which contain all multi-qubit connections defined by the factories present in the config.- Type
set
-
property
factories
¶ The tuple of
GateFactory
s which define all gates present in this config.- Type
tuple
-
property
dim
¶ The dimension of systems (e.g., 2 for qubits, 3 for qutrits).
- Type
int
-
static
from_parameterized
(multiqubit_fact, parameters, mode='ZXZ')¶ A convenience method for making a configuration file for a quantum system with fixed multi-qubit gates with nonhomogenous parameters.
The example below defines a config containing two fixed rotation entangling gates which are specific parameter combinations of the defined
trueq.config.GateFactory
. This enables a config to be defined using known interactions, such as the Cross-Resonance interaction, but have unique fixed parameter gate for each physical gate present in hardware.import trueq as tq iSwapLike_mat = [[1, 0, 0, 0], [0, 0, -1j, 0], [0, -1j, 0, 0], [0, 0, 0, "exp(-1j * phi / 180)"], ] iSwapLike_factory = tq.config.GateFactory.from_matrix( "iSwapLike", iSwapLike_mat ) parameters = {(0, 1): 90, (2, 3): 45} conf = tq.Config.from_parameterized(iSwapLike_factory, parameters) conf.iSwapLike_0_1()
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".- Name:
-
- iSwapLike_0_1()
- Likeness:
-
- Non-Clifford
- Generators:
-
- 'IZ': -14.324
- 'XX': 90.0
- 'YY': 90.0
- 'ZI': -14.324
- 'ZZ': 14.324
- Matrix:
-
- Parameters
multiqubit_fact (
GateFactory
) – A gate factory that can generate the nonhomogenous gates that can be implemented in the system.parameters (
dict
) – A dictionary mapping tuples of qubit labels to parameters of the given mulit-qubit factory to be used on those particular qubits.mode (
str
) – The mode used to decompose single-qubit operations.
- Return type
- Raises
ValueError – If an incorrect number of labels are provided in
parameters
.ValueError – If the provided
mode
is not defined inQubitMode
.
-
static
from_yaml
(conf)¶ Construct a
Config
from the contents of a yaml file, or filename.A minimal example
.yaml
file is presented below. The YAML below defines a four-qubit system, where the gates available are:Arbitrary \(X(\phi)\) rotations, which cannot be applied while the neighboring qubits are in use.
Arbitrary \(Z(\phi)\) rotations, which can be applied in parallel with anything
CNOT, which can never be used in parallel with any other gate.
Note that gates can either be defined by the
Hamiltonian
orMatrix
keyword, but never both.import trueq as tq file_contents = ''' Dimension: 2 Mode: ZXZXZ Gates: - X(phi=90): Hamiltonian: - ['X', phi] Involving: (0,) : (1, 2) (1,) : (0, 2) (2,) : (1, 3) (3,) : (2,) - Z: Hamiltonian: - ['Z', phi] Involving: (0,) : () (1,) : () (2,) : () (3,) : () - U3(theta, phi, lam): Hamiltonian: - [Z, 180 / pi * phi] - [Y, 180 / pi * theta] - [Z, 180 / pi * lam] - CNOT: Matrix: - [1, 0, 0, 0] - [0, 1, 0, 0] - [0, 0, 0, 1] - [0, 0, 1, 0] Involving: (0, 1) : (2, 3) (1, 2) : (0, 3) (2, 3) : (0, 1)''' # Configs can be loaded from either a filename, or the contents of a file: config = tq.Config.from_yaml(file_contents) config
Mode: ZXZXZ Dimension: 2 Gates: - X(phi=90.0): Hamiltonian: - [X, phi] Involving: {'(0,)': '(1, 2)', '(1,)': '(0, 2)', '(2,)': '(1, 3)', '(3,)': '(2,)'} - Z(phi): Hamiltonian: - [Z, phi] Involving: {'(0,)': (), '(1,)': (), '(2,)': (), '(3,)': ()} - U3(theta, phi, lam): Hamiltonian: - [Z, 57.29577951308235*phi] - [Y, 57.29577951308232*theta] - [Z, 57.29577951308235*lam] - CNOT: 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: {'(0, 1)': '(2, 3)', '(1, 2)': '(0, 3)', '(2, 3)': '(0, 1)'}
The above config contains information about physical limitations of the gates. This information is encoded in the
Involving
key inside of the gate definitions. This involving key is a dictionary whose keys are the sets of systems on which the gate can act. In the YAML example above, theX
gate can be applied to any of qubits0, 1, 2, 3
.The values of the dictionary list which systems cannot have any gates applied to them while the gate acts on the systems specified by the key. For example, when the
X
gate in the above config is run on qubit0
, no other gates may be applied in parallel on either qubit1
or2
.This
Involving
information is used during compilation to ensure that circuits are compatable with the physical limitations of the hardware.Since we allow arbitrary functional definition of gates using functions that can be using
parse_func()
, theConfig
object can be used to generate newNativeGate
objects on the fly based upon the variables defined in the YAML.In the above YAML file, the
U3
gate specifies the(theta, phi, lam)
parameters. If these parameters are not provided manually they will automatically be found. The manual specification above is to enforce a specific parameter order when calling the gate factory without using kwargs, this can be seen in the following example:assert config.U3(2, 3, 4) == config.U3(theta=2, phi=3, lam=4)
Parameters which have been set, such as the
X(phi=90)
gate above, will be set in theGateFactory
during construction and cannot be altered. AllNativeGate
s built from these definitions will have the fixed parameter present in their parameter list. SeeGateFactory
for more information.assert config.X() == tq.Gate.sx assert config.X().parameters == {"phi": 90} config.X()
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".- Name:
-
- X(phi)
- Aliases:
-
- Gate.sx
- Gate.cliff5
- Parameters:
-
- phi = 90.0
- Generators:
-
- 'X': 90.0
- Matrix:
-
- Parameters
conf (
str
) – Path to a configuration YAML file, or a YAML file as a string.- Raises
ValueError – If
conf
is not a string to a YAML file or the contents of a valid YAML file.
-
property
mode
¶ The mode by which arbitrary single-qubit gates are decomposed in the system to native gates.
- Return type
str
-
subset
(labels=None, names=None, strict=True)¶ Creates a new
Config
object containing a subset of the gate factories present in this configuration, and each of whoseinvolving
rules has been trimmed based on the givenlabels
.import trueq as tq # create an basic config using cz gates as the entangler config = tq.Config.basic(entangler=tq.Gate.cz) # make a new config acting only on qubits 2, 3 restricted to gates z, cz config.subset(labels=[2, 3], names=["z", "cz"])
Mode: ZXZXZ Dimension: 2 Gates: - z(phi): Hamiltonian: - [Z, phi] - cz: Matrix: - [1.0, 0.0, 0.0, 0.0] - [0.0, 1.0, 0.0, 0.0] - [0.0, 0.0, 1.0, 0.0] - [0.0, 0.0, 0.0, -1.0]
Note
This preserves the ordering of factories as specified by
names
kwarg, if nonames
are specified, then this preserves the ordering of factories present infactories
.- Parameters
labels (
Iterable
) – The labels to be present in the new config. IfNone
, no filtering based on labels is done.names (
Iterable
) – A list of gate factory names to keep. IfNone
, no filtering based on factory names is done.strict (
bool
) – Whether to filter gate factories based on being a strict subset of the givenlabels
(True
), or based on overlapping withlabels
at all (False
).
- Return type
-
property
factory_names
¶ A list of all factory names present in the config.
- Type
list
-
get
(name, labels=None)¶ Returns a list of factories which match the specified name and optionally match the required labels.
import trueq as tq config = tq.Config.basic(entangler=tq.Gate.cz) assert config.get('cz') == [config.cz] assert config.get('cz', (0, 1)) == [config.cz]
- Parameters
name (
str
) – The name of a factory.labels (
tuple
|int
|None
) – Optional, the labels on which the factory operates.
- Return type
list
- Raises
KeyError – If the specified factory cannot be found.
Gate Factory¶
-
class
trueq.config.
GateFactory
(name, layers=None, involving=None, matrix=None, hamiltonian=None, parameters=None, sys_dim=2)¶ Constructs new
NativeGate
s according to a functional description.This is useful when you have a description of a gate such as \(X(\theta)\) which is an arbitrary rotation around the
X
axis by some angle \(\theta\). Convenience methods are provided which enable construction of the factory using either a unitary matrix definition (GateFactory.from_matrix()
), Hamiltonian description (GateFactory.from_hamiltonian()
), or through a function (GateFactory.from_function()
).Examples:
from trueq.config import GateFactory y_factory = GateFactory.from_hamiltonian('y', [['Y', 'phi']]) y_factory(phi=90)
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".- Name:
-
- y(phi)
- Aliases:
-
- Gate.sy
- Gate.cliff7
- Parameters:
-
- phi = 90
- Generators:
-
- 'Y': 90.0
- Matrix:
-
from trueq.config import GateFactory iSwapLike_mat = [[1, 0, 0, 0], [0, 0,-1j, 0], [0,-1j, 0, 0], [0, 0, 0, "exp(-1j * phi / 180)"]] iSwap_factory = GateFactory.from_matrix("iSwapLike", iSwapLike_mat) iSwap_factory(phi=45)
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".- Name:
-
- iSwapLike(phi)
- Likeness:
-
- Non-Clifford
- Parameters:
-
- phi = 45
- Generators:
-
- 'IZ': -7.162
- 'XX': 90.0
- 'YY': 90.0
- 'ZI': -7.162
- 'ZZ': 7.162
- Matrix:
-
from trueq.config import GateFactory import numpy as np u3_factory = GateFactory.from_hamiltonian( "U3Gate", [ ["Z", "phi * 180 / pi"], ["Y", "theta * 180 / pi"], ["Z", "lam * 180 / pi"], ], parameters=["theta", "phi", "lam"], ) u3_factory(np.pi/2, 0, 0)
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".- Name:
-
- U3Gate(theta, phi, ...)
- Aliases:
-
- Gate.sy
- Gate.cliff7
- Parameters:
-
- theta = 1.570796
- phi = 0
- lam = 0
- Generators:
-
- 'Y': 90.0
- Matrix:
-
The
U3
gate as defined by IBM, in this case the gate is defined in terms of radians in the IBM definition (note that this definition is in multiplicative order, but the above Hamiltonian definition is in time order):\(U3(\theta, \phi, \lambda) = Z(\lambda)Y(\theta)Z(\phi)\)
This radian based definition means that there has to be a conversion constant on each of the parameters
phi, theta, lam
from radians to degrees. Additionally, the order of the parameters is important in theU3
definition, so theGateFactory
is defined with theparameters
options so that the order is recorded correctly.- Parameters
name (
str
) – Name of the gate operation which must be a valid identifier examples:CZ
,X
,CNOT
.layers (
list
) – A list ofLayer
objects which define the factory. When theGateFactory
is called, the provided parameters are passed to each layer, and all layers are multiplied together in order to construct aNativeGate
.involving (
dict
) – A dictionary describing the limitations of the gate, seeConfig
description for a more in-depth description.parameters (
list
|dict
|None
) –An optional specification of the parameters present, this accepts the following inputs:
None
Parameter names are inferred from the layers in a deterministic but arbitrary order, and all parameters are free to be set to any value.list
A list of names, this specifies the preferred ordering of the parameters when calling the factory withargs
. Names present in this list must match the parameters present in the layers exactly, and all parameters are free to be set to any value.dict
The keys of the dictionary specify the parameters used in the factory, the values may either beNone
signifying that the parameter may be set to anything, or a specific value that the parameter is always set to. Any parameter with a fixed value will always be included in the constructedNativeGate
.
sys_dim (
int
) – The dimension of the subsystem defined by this gate, defaults to2
for qubits.
- Raises
ValueError – If invalid input arguments or argument combinations are provided.
-
property
layers
¶ The
Layer
s which make up theGateFactory
.During construction of the
NativeGate
these layers construct matrices which are multiplied together in order, the resultant matrix is the unitary representation of the gate.- Type
list
-
classmethod
from_function
(name, function, involving=None, parameters=None, sys_dim=2)¶ Constructs a
GateFactory
from a function which returns a square, unitary matrix.The matrix output of the function must obey the following conditions:
Square and unitary.
The matrix can be written as a product of
n
Rotation
s with a single optionalFixedRotation
.Each parameter accepted by the function can be mapped to exactly one
Rotation
.This means there can be no more than
n+1
total layers, wheren
is the number of arguments of the function.
If these conditions are met, the
Layer
s which define the function are used to construct theGateFactory
.- Parameters
name (
str
) – Name of the gate operation which must be a valid identifier, for example:CZ
,X
,CNOT
.function (
function
) – A function that, when called, generates a unitary matrix.involving (
dict
) – A dictionary describing the limitations of the gate, seeConfig
description for a more in-depth description.parameters (
list
|dict
|None
) – An optional specification of the parameters present, seeGateFactory
for more details.sys_dim (
int
) – The dimension of the subsystem defined by this gate, defaults to2
for qubits.
- Raises
ValueError – If invalid input arguments or argument combinations are provided.
-
classmethod
from_matrix
(name, matrix, involving=None, parameters=None, sys_dim=2)¶ Constructs a
GateFactory
from a matrix definition which is square and unitary.The matrix must obey the following conditions:
Square and unitary.
The matrix can be written as a product of
n
Rotation
s with a single optionalFixedRotation
.Each parameter present in the matrix can be mapped to exactly one
Rotation
.This means there can be no more than
n+1
total layers, wheren
is the number of free parameters in the matrix.
- Parameters
name (
str
) – Name of the gate operation which must be a valid identifier examples:CZ
,X
,CNOT
.matrix (
list
) – A unitary matrix description of the gate, this may include string definitions of functions which will be parsed on instantiation and converted into numerical representations which are equivalent to the Hamiltonian description. The matrix definition must be writable as a product of generator matrices \(\prod_i e^{i A_i \cdot t_i}\), where each parameter \(t_i\) appear only once. If this requirement is too restrictive, the gate may be defined more generally using the Hamiltonian.involving (
dict
) – A dictionary describing the limitations of the gate, seeConfig
description for a more in-depth description.parameters (
list
|dict
|None
) – An optional specification of the parameters present, seeGateFactory
for more details.sys_dim (
int
) – The dimension of the subsystem defined by this gate, defaults to2
for qubits.
- Raises
ValueError – If invalid input arguments or argument combinations are provided.
-
classmethod
from_hamiltonian
(name, hamiltonian, involving=None, parameters=None, sys_dim=2)¶ Constructs a
GateFactory
from a Hamiltonian definition.Example:
from trueq.config import GateFactory import numpy as np u3_factory = GateFactory.from_hamiltonian( "U3Gate", [ ["Z", "phi * 180 / pi"], ["Y", "theta * 180 / pi"], ["Z", "lam * 180 / pi"], ], parameters=["theta", "phi", "lam"], ) u3_factory(np.pi/2, 0, 0)
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".- Name:
-
- U3Gate(theta, phi, ...)
- Aliases:
-
- Gate.sy
- Gate.cliff7
- Parameters:
-
- theta = 1.570796
- phi = 0
- lam = 0
- Generators:
-
- 'Y': 90.0
- Matrix:
-
The
U3
gate as defined by IBM, in this case the gate is defined in terms of radians in the IBM definition (note that this definition is in multiplicative order, but the above Hamiltonian definition is in time order):\(U3(\theta, \phi, \lambda) = Z(\lambda)Y(\theta)Z(\phi)\)
- Parameters
name (
str
) – Name of the gate operation which must be a valid identifier examples:CZ
,X
,CNOT
.hamiltonian (
list
) – A Hamiltonian generator description of the gate, defined above. The values may be strings which are parsable into values. The list of provided Hamiltonians will be applied in time ordering, and are not required to commute.involving (
dict
) – A dictionary describing the limitations of the gate, seeConfig
description for a more in-depth description.parameters (
list
|dict
|None
) – An optional specification of the parameters present, seeGateFactory
for more details.sys_dim (
int
) – The dimension of the subsystem defined by this gate, defaults to2
for qubits.
- Raises
ValueError – If invalid input arguments or argument combinations are provided.
-
subset
(labels=None, strict=True)¶ Creates a new
GateFactory
object with a set ofinvolving
rules that is a subset of this factory’s rules.import trueq as tq import numpy as np involving = {(0, 1): (2, 3), (1, 2): (3,), (2, 3): (1,), (3, 8): ()} factory = tq.config.GateFactory.from_matrix( "cz", np.diag([1, 1, 1, -1]), involving=involving ) subset = factory.subset(labels=(0, 1, 2), strict=True) assert subset.involving.rules == {(0, 1): (2, 3), (1, 2): (3,)} subset
GateFactory(name='cz', layers=[FixedRotation(<matrix>)], involving=InvolvingRule(rules={(0, 1): (2, 3), (1, 2): (3,)}, default_allow=False))
- Parameters
labels (
Iterable
) – The labels to be present in the new config. IfNone
, no filtering based on labels is done.strict (
bool
) – Whether to filterinvolving
labels based on being a strict subset of the givenlabels
(True
), or based on overlapping withlabels
at all (False
).
- Return type
-
fix_parameters
(involving=None, **params)¶ Constructs a new
GateFactory
with an updatedparameters
dictionary, and optionally a new set ofinvolving
restrictions.import trueq as tq factory = tq.config.GateFactory.from_hamiltonian("x", [['X','phi']]) factories = [factory.fix_parameters(phi=x) for x in [0, 90, 180, 270]] tq.Config(factories=factories)
Mode: ZXZXZ Dimension: 2 Gates: - x(phi=0): Hamiltonian: - [X, phi] - x(phi=90): Hamiltonian: - [X, phi] - x(phi=180): Hamiltonian: - [X, phi] - x(phi=270): Hamiltonian: - [X, phi]
- Parameters
involving (
dict
) – A dictionary describing the limitations of the gate, seeConfig
description for a more in-depth description. If this isNone
, use the exisinginvolving
, otherwise overwrite.params (
dict
) –kwargs
which specify the parameter values to fix, see the example above.
- Return type
-
property
pauli_const
¶ If this factory is a simple rotation about a single Pauli axis, specifically a gate which is representable by \(exp^{-i \frac{\pi}{360} \text{pauli} \cdot \text{constant} \cdot \text{param}}\), then
pauli_const
is a tuple which represents the rotation,(Pauli, constant)
.If the gate cannot be written in terms of a single Pauli rotation, then
pauli_const
is(None, None)
.from trueq.config import GateFactory import numpy as np factory = GateFactory.from_hamiltonian("Rz", [["Z", "5 * phi"]]) assert factory.pauli_const[0] == 'Z' assert abs(factory.pauli_const[1] - 5) < 1e-10
- Type
tuple
-
property
pauli_angle
¶ If this factory is a simple fixed angle rotation about a single Pauli axis which may be written as \(exp^{-i \frac{\pi}{360} \cdot \text{Pauli} \cdot \text{angle}}\), then
pauli_angle
is a tuple which represents the rotation,(Pauli, angle)
.If the gate cannot be written in terms of a single Pauli rotation, then
pauli_angle
is(None, None)
.from trueq.config import GateFactory factory = GateFactory.from_hamiltonian("Rz", [['Z', '90']]) assert factory.pauli_angle == ('Z', 90)
- Type
tuple
-
make_gate
(*args, **kwargs)¶ Returns a
NativeGate
based upon the parameters provided.Note
Constructed
NativeGate
objects are memoized and cached based upon thekwargs
.- Return type
- Raises
ValueError – If the arguments do not match the parameters found in the description of the gate.
KeyError – If a functional parameter is missing from
kwargs
.
-
property
n_sys
¶ The number of systems this gate factory acts on.
- Type
int
-
property
sys_dim
¶ The dimension of the subsystems this gate factory acts on.
- Type
int
-
property
width
¶ The width of gates produced by this factory, e.g., 2 for a single qubit, 4 for two qubits, etc.
- Type
int
-
property
parameters
¶ The dictionary of all the parameters in the functional definition of the gate.
The values of the dictionary contain either the fixed value associated with that paramter, or
None
signifying that any value is allowed.- Type
dict
-
property
free_parameters
¶ A
list
containing all of the free parameters.- Type
list
-
property
n_params
¶ The number of free parameters in the functional definition of the gate.
- Type
int
-
property
fixed_parameters
¶ Fixed values in the factory which cannot be changed, these values are added to all parameters present in
NativeGate
s constructed by this factory.- Type
dict
-
property
involving
¶ A specification of which systems can be involved in the gate operation.
- Type
-
to_dict
()¶ Returns a dictionary representation of the factory object.
- Return type
dict
-
static
from_dict
(dic)¶ Create a
GateFactory
from a dictionary representation.- Parameters
dic – A dictionary representation of the factory as generated by
GateFactory.to_dict()
.- Return type
-
property
is_u1
¶ Tests if a
trueq.config.GateFactory
is aU1
gate. In order to pass this test, the factory must have the same parameter names and same definition as theU1Gate
defined in Qiskit.
-
property
is_u2
¶ Tests if a
trueq.config.GateFactory
is aU2
gate. In order to pass this test, the factory must have the same parameter names and same definition as theU2Gate
defined in Qiskit.
-
property
is_u3
¶ Tests if a
trueq.config.GateFactory
is aU3
gate. In order to pass this test, the factory must have the same parameter names and same definition as theU3Gate
defined in Qiskit.
-
property
is_r
¶ Tests if a
trueq.config.GateFactory
is anR
gate. An R gate is defined as a \(Z(-\theta) X(90) Z(\theta)\) gate in multiplication ordering.
Involving Rule¶
-
class
trueq.config.
InvolvingRule
(rules=None, default_allow=None)¶ A dictionary containing allowable sets of systems.
from trueq.config import InvolvingRule # In the default case, any involving check returns as valid rule = InvolvingRule() (5, ) in rule, rule[(1, 5)]
(True, ())
from trueq.config import InvolvingRule # In the case where rules are provided, it will return the rules # if asked specifically about the items, and nothing otherwise rule = InvolvingRule(rules={(0,): (2, 3, 4)}) rule[(0,)], (1, ) in rule
((2, 3, 4), False)
from trueq.config import InvolvingRule # In the case where rules are provided, and `default_allow=True` # it will return if asked specifically about the contained items, # and will return `True` for all other cases rule = InvolvingRule(rules={(0,): (2, 3, 4)}, default_allow=True) rule[(0,)], rule[(1,)]
((2, 3, 4), ())
See
Config
for more details on how these rules are used.- Parameters
rules (
dict
) – A dictionary with both keys and values being tuples of ints. The key represents allowed operations, and the values represent blocking operations that get placed on other qubits due to the key labels.default_allow (
bool
|None
) – Defines the behavior if a key is not present in the ruleset. IfTrue
, then if a key is not found in the rules,InvolvingRule
acts as though it were present. IfFalse
,InvolvingRule
raiseKeyErrors
. IfNone
, thendefault_allow
is set toTrue
if no rules are provided, andFalse
if rules are provided.
-
keys
()¶ Returns the keys for given rules.
- Return type
generator
-
values
()¶ Returns the values for given rules.
- Return type
generator
-
items
()¶ Returns the key, value pairs for given rules.
- Return type
generator
-
unordered_get
(key)¶ Gets the value associated with a given key, regardless of the ordering.
This returns a list of tuples, where the tuples contain the required ordering as defined by the rules, along with the value associated with that key.
from trueq.config import InvolvingRule rule = InvolvingRule(rules={(0, 1): (2, 3, 4)}) rule.unordered_get((1, 0))
[((0, 1), (2, 3, 4))]
This is useful when connectivity matters, but not the strict ordering of labels.
- Parameters
key (
tuple
) – The key to fetch, must be a tuple of ints.- Return type
tuple
- Raises
KeyError – If labels are not present in rules.
-
unordered_contains
(key)¶ Determines if any permutation of the given key is present in rules.
from trueq.config import InvolvingRule rule = InvolvingRule(rules={(0, 1): (2, 3, 4)}) rule.unordered_contains((1, 0))
True
This is useful when connectivity matters, but not the strict ordering of labels.
- Parameters
key (
tuple
) – The key to fetch, must be a tuple of ints.- Return type
bool
Parsing functions¶
-
trueq.config.factory.
parse_func
(func_str)¶ Creates a Python function from a string containing numbers, variables, and standard math functions.
from trueq.config.factory import parse_func, KNOWN_FUNCTIONS f = parse_func('5*sin(x) + y') print(f(x=np.pi, y=1)) # all functions in KNOWN_FUNCTIONS can be parsed print(KNOWN_FUNCTIONS.keys())
1.0000000000000007 dict_keys(['sin', 'cos', 'tan', 'exp', 'sqrt', 'arcsin', 'arccos', 'arctan', 'round', 'ceil', 'floor', 'int', 'float', 'complex', 'pi', 'list', 'tuple', 'dict', 'set'])
Note
Parsing functions adds an overhead relative to directly calling numpy functions. The overhead is approximately a factor of 20 for simple functions.
- Parameters
func_str (
str
) – A string describing the function to be created.- Return type
func
-
trueq.config.factory.
find_parameters
(func_str)¶ Finds all free parameters inside the text description of a fuction using exception handling.
from trueq.config.factory import find_parameters params = find_parameters("sin(x) + phi / 2 + theta") assert set(params) == {"x", "phi", "theta"}
- Parameters
func_str (
str
) – A string describing the function.- Return type
list
Pre-Defined factories¶
-
trueq.config.factory.
u1_factory
= GateFactory(name='U1Gate', layers=[Rotation.from_pauli('Z', 'theta', 57.29577951308235)], parameters={'theta': None})¶ A single qubit
U1
Gate as defined by Qiskit, equivalent to a \(Z(\theta)\) rotation.
-
trueq.config.factory.
u2_factory
= GateFactory(name='U2Gate', layers=[Rotation.from_pauli('Z', 'phi', 57.29577951308235), FixedRotation.from_pauli('Y', 90.0), Rotation.from_pauli('Z', 'lam', 57.29577951308235)], parameters={'phi': None, 'lam': None})¶ A single qubit
U2
Gate as defined by Qiskit, equivalent to a \(Z(\lambda)Y(90)Z(\phi)\) rotation.
-
trueq.config.factory.
u3_factory
= GateFactory(name='U3Gate', layers=[Rotation.from_pauli('Z', 'phi', 57.29577951308235), Rotation.from_pauli('Y', 'theta', 57.29577951308232), Rotation.from_pauli('Z', 'lam', 57.29577951308235)], parameters={'theta': None, 'phi': None, 'lam': None})¶ A single qubit
U3
Gate as defined by Qiskit, equivalent to a \(Z(\lambda)Y(\theta)Z(\phi)\) rotation.
-
trueq.config.factory.
phasedXPow_factory
= GateFactory(name='PhasedXPow', layers=[Rotation.from_pauli('Z', 'lam', 180.0), Rotation.from_pauli('X', 'theta', 180.0), Rotation.from_pauli('Z', 'lam', -180.0)], parameters={'lam': None, 'theta': None})¶ A single qubit
PhasedXPow
Gate as defined by Cirq, equivalent to a \(Z(-\lambda / 180)X(\theta / 180)Z(\lambda / 180)\) rotation.
-
trueq.config.factory.
r_factory
= GateFactory(name='R', layers=[Rotation.from_pauli('Z', 'theta', -1.0), FixedRotation.from_pauli('X', 90.0), Rotation.from_pauli('Z', 'theta', 1.0)], parameters={'theta': None})¶ A single qubit
R
Gate, equivalent to a \(Z(-\theta / 180)X(90)Z(\theta / 180)\) rotation.
-
trueq.config.factory.
ms_factory
= GateFactory(name='MS', layers=[Rotation.from_pauli('XX', 'theta', 1.0)], parameters={'theta': None})¶ A two qubit
Molmer-Sorensen
Gate, equivalent to a \(XX(\theta)\) rotation.
-
trueq.config.factory.
fsim_factory
= GateFactory(name='FSim', layers=[Rotation(<matrix>, 'phi'), Rotation(<matrix>, 'theta')], parameters={'theta': None, 'phi': None})¶ A two qubit
F-Sim
Gate as defined by Cirq.