qml.Select¶
- class Select(ops, control, work_wires=None, id=None)[source]¶
Bases:
Operation
The
Select
operator, also known as multiplexer or multiplexed operation, applies different operations depending on the state of designated control wires.Select|i⟩⊗|ψ⟩=|i⟩⊗Ui|ψ⟩If the applied operations {Ui} are all single-qubit Pauli rotations about the same axis, with the angle determined by the control wires, this is also called a uniformly controlled rotation gate.
See also
- Parameters:
ops (list[Operator]) – operations to apply
control (Sequence[int]) – the wires controlling which operation is applied. At least ⌈log2K⌉ wires are required for K operations.
work_wires (Union[Wires, Sequence[int], or int]) – auxiliary wire(s) that may be utilized during the decomposition of the operator into native operations. For details, see the section on the unary iterator decomposition below.
id (str or None) – String representing the operation (optional)
Note
The position of the operation in the list determines which qubit state implements that operation. For example, when the qubit register is in the state |00⟩, we will apply
ops[0]
. When the qubit register is in the state |10⟩, we will applyops[2]
. To obtain the list positionindex
for a given binary bitstring representing the control state we can use the following relationship:index = int(state_string, 2)
. For example,2 = int('10', 2)
.Example
>>> dev = qml.device('default.qubit', wires=4) >>> ops = [qml.X(2), qml.X(3), qml.Y(2), qml.SWAP([2,3])] >>> @qml.qnode(dev) >>> def circuit(): >>> qml.Select(ops, control=[0,1]) >>> return qml.state() ... >>> print(qml.draw(circuit, level='device')()) 0: ─╭○─╭○─╭●─╭●────┤ State 1: ─├○─├●─├○─├●────┤ State 2: ─╰X─│──╰Y─├SWAP─┤ State 3: ────╰X────╰SWAP─┤ State
Unary iterator decomposition
Generically,
Select
is decomposed into one multi-controlled operator for each target operator. However, if auxiliary wires are available, a decomposition using a “unary iterator” can be applied. It was introduced by Babbush et al. (2018).Principle
Unary iteration leverages auxiliary wires to store intermediate values for reuse between the different multi-controlled operators, avoiding unnecessary recomputation. In addition to this caching functionality, unary iteration reduces the cost of the computation directly, because the involved reversible AND (or Toffoli) gates can be implemented at lower cost if the target is known to be in the |0⟩ state (see
TemporaryAND
).For K operators to be Select-applied, c=⌈log2K⌉ control wires are required. Unary iteration demands an additional c−1 auxiliary wires. Below we first show an example for K being a power of two, i.e., K=2c. Then we elaborate on implementation details for the case K<2c, which we call a partial Select operator.
Example
Assume that we want to Select-apply K=8=23 operators to two target wires, which requires c=⌈log2K⌉=3 control wires. The generic decomposition for this takes the form
0: ─╭○─────╭○─────╭○─────╭○─────╭●─────╭●─────╭●─────╭●─────┤ 1: ─├○─────├○─────├●─────├●─────├○─────├○─────├●─────├●─────┤ 2: ─├○─────├●─────├○─────├●─────├○─────├●─────├○─────├●─────┤ 3: ─├U(M0)─├U(M1)─├U(M2)─├U(M3)─├U(M4)─├U(M5)─├U(M6)─├U(M7)─┤ 4: ─╰U(M0)─╰U(M1)─╰U(M2)─╰U(M3)─╰U(M4)─╰U(M5)─╰U(M6)─╰U(M7)─┤.
Unary iteration then uses c−1=2 auxiliary wires, denoted
aux0
andaux1
below, to first rewrite the control structure:0: ─╭○───────○╮─╭○───────○╮─╭○───────○╮─╭○───────○╮─╭●───────●╮─╭●───────●╮─╭●───────●╮─╭●───────●╮─┤ 1: ─├○───────○┤─├○───────○┤─├●───────●┤─├●───────●┤─├○───────○┤─├○───────○┤─├●───────●┤─├●───────●┤─┤ aux0: ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ ╰─╭●───●╮─╯ │ 2: ───├○───○┤─────├●───●┤─────├○───○┤─────├●───●┤─────├○───○┤─────├●───●┤─────├○───○┤─────├●───●┤───┤ aux1: ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ ╰─╭●──╯ │ 3: ─────├U(M0)──────├U(M1)──────├U(M2)──────├U(M3)──────├U(M4)──────├U(M5)──────├U(M6)──────├U(M7)──┤ 4: ─────╰U(M0)──────╰U(M1)──────╰U(M2)──────╰U(M3)──────╰U(M4)──────╰U(M5)──────╰U(M6)──────╰U(M7)──┤
Here, we used the symbols
0: ─╭●── ─●─╮─ 1: ─├●── and ─●─┤─ 2: ╰─── ───╯
for
TemporaryAND
and its adjoint, respectively, and skipped drawing the auxiliary wires in areas where they are guaranteed to be in the state |0⟩. We will need three simplification rules for pairs ofTemporaryAND
gates:─○─╮─╭○── ── ─○─╮─╭○── ─╭○─ ─○─╮─╭●── ─╭●──── ─○─┤─├○── = ──, ─○─┤─├●── = ─│──, and ─●─┤─├○── = ─│──╭●─. ───╯ ╰─── ── ───╯ ╰─── ─╰X─ ───╯ ╰─── ─╰X─╰X─
Applying these simplifications reduces the computational cost of the
Select
template:0: ─╭○────────────────╭○──────────────────╭●─────────────────────╭●─────────────────●╮─┤ 1: ─├○────────────────│───────────────────│──╭●──────────────────│──────────────────●┤─┤ aux0: ╰─╭●─────╭●────●╮─╰X─╭●─────╭●─────●╮─╰X─╰X─╭●─────╭●─────●╮─╰X─╭●─────╭●─────●╮─╯ │ 2: ───├○─────│─────●┤────├○─────│──────●┤───────├○─────│──────●┤────├○─────│──────●┤───┤ aux1: ╰─╭●───╰X─╭●──╯ ╰─╭●───╰X──╭●──╯ ╰─╭●───╰X──╭●──╯ ╰─╭●───╰X──╭●──╯ │ 3: ─────├U(M0)──├U(M1)─────├U(M2)───├U(M3)────────├U(M4)───├U(M5)─────├U(M6)───├U(M7)──┤ 4: ─────╰U(M0)──╰U(M1)─────╰U(M2)───╰U(M3)────────╰U(M4)───╰U(M5)─────╰U(M6)───╰U(M7)──┤
An additional cost reduction then results from the fact that the
TemporaryAND
gate and its adjoint require four and zeroT
gates, respectively, in contrast to the sevenT
gates required by a decomposition ofToffoli
.For general c and K=2c, the decomposition takes a similar form, with alternating control and auxiliary wires.
An implementation of the unary iterator is achieved in the following steps: We first define a recursive sub-circuit
R
; given L operators and 2⌈log2(L)⌉+1 control and auxiliary wires, there are three cases thatR
distinguishes. First, ifL>1
, it applies the circuitaux_j: ╭R ─╭●────╭●────●─╮─ j+1: ├R = ─├○────│─────●─┤─ aux_j+1: ╰R ╰──R──╰X─R────╯ ,
where each label
R
symbolizes a call toR
itself, on the next recursion level. These next-level calls use L′=2⌈log2(L)⌉−1 (i.e. half of L, rounded up to the next power of two) and L−L′ (i.e. the rest) operators, respectively.Second, if
L=1
, the single operator is applied, controlled on the first control wire. Finally, ifL=0
,R
does not apply any operators.With
R
defined, we are ready to outline the main circuit structure:Apply the left-most
TemporaryAND
controlled on qubits0
and1
.Split the target operators into four “quarters” (often with varying sizes) and apply the first quarter using
R
.Apply
[X(0), CNOT([0, "aux0"]), X(0)]
.Apply the second quarter using
R
.Apply
[CNOT([0, "aux0"]), CNOT([1, "aux0"])]
.Apply the third quarter using
R
.Apply
[CNOT([0, "aux0"])]
.Apply the last quarter using
R
.Apply the right-most
adjoint(TemporaryAND)
controlled on qubits0
and1
.
Partial Select decomposition
The unary iterator decomposition of the
Select
template can be simplified further if both of the following criteria are met:There are fewer target operators than would maximally be possible for the given number of control wires, i.e. K<2c.
The state |ψ⟩ of the control wires satisfies ⟨j|ψ⟩=0 for all computational basis states with j≥K.
We do not derive this reduction here but discuss the modifications to the implementation above that result from it.
Given K=2c−b operators, where c is defined as above and we have 0≤b<2c−1, the nine steps above are modified into one of three variants. In each variant, the first 2c−1 operators are applied in two equal portions, containing 2c−2 operators each. After this, ℓ=2c−1−b operators remain and the three circuit variants are distinguished, based on ℓ:
if ℓ≥2c−2, the following, rather generic, circuit is applied:
0: ─╭○─────╭○─────╭●────────╭●─────●─╮─ 1: ─├○─────│──────│──╭●─────│──────●─┤─ aux0: ╰──╭R──╰X─╭R──╰X─╰X─╭R──╰X─╭R────╯ 2: ────├R─────├R────────├R─────├R────── aux1: ╰R ╰R ╰R ╰R .
Here, each operator with three
R
labels symbolizes a call toR
. The first call in the second half applies 2⌈log2(ℓ)⌉−1 operators. Note that this case is triggered if K is larger than or equal to 34 of the maximal capacity for c control wires. Also note how the two middleTemporaryAND
gates were merged into two CNOTs, like for the non-partial Select operator.if 1<ℓ<2c−2, the following circuit is applied:
0: ─╭○─────╭○─────○─╮╭●─────╭●─────●─╮─ 1: ─├○─────│──────●─┤│──────│────────│─ aux0: ╰──╭R──╰X─╭R────╯│ │ │ 2: ────├R─────├R─────├○─────│──────●─┤─ aux1: ╰R ╰R ╰───R──╰X──R────╯
where the second half may skip more than one control and auxiliary wire each. In this diagram, both the operators with three and one
R
labels represent calls toR
, with single-label instances applying fewer operators. The first call toR
in the second half applies 2⌈log2(ℓ)⌉−1 operators. The middle elbows act on distinct wire triples and can not be merged as above.if ℓ=1, the following circuit is applied:
0: ─╭○─────╭○─────○─╮╭●── 1: ─├○─────│──────●─┤│─── aux0: ╰──╭R──╰X─╭R────╯│─── 2: ────├R─────├R─────│─── aux1: ╰R ╰R ╰U .
Here, the three connected
R
labels symbolize a call toR
and apply 2c−2 operators each. The controlled gate on the right applies the single remaining operator.
Attributes
Arithmetic depth of the operator.
The basis of an operation, or for controlled gates, of the target operation.
Batch size of the operator if it is used with broadcasted parameters.
The control wires.
Control wires of the operator.
Create data property
Gradient computation method.
Gradient recipe for the parameter-shift method.
Integer hash that uniquely represents the operator.
Dictionary of non-trainable variables that this operation depends on.
Custom string to label a specific operator instance.
This property determines if an operator is hermitian.
String for the name of the operator.
Number of dimensions per trainable parameter of the operator.
Number of trainable parameters that the operator depends on.
Number of wires the operator acts on.
Operations to be applied.
Returns the frequencies for each operator parameter with respect to an expectation value of the form ⟨ψ|U(p)†ˆOU(p)|ψ⟩.
Trainable parameters that the operator depends on.
A
PauliSentence
representation of the Operator, orNone
if it doesn't have one.A dictionary containing the minimal information needed to compute a resource estimate of the operator's decomposition.
The wires of the target operators.
All wires involved in the operation.
- arithmetic_depth¶
Arithmetic depth of the operator.
- basis¶
The basis of an operation, or for controlled gates, of the target operation. If not
None
, should take a value of"X"
,"Y"
, or"Z"
.For example,
X
andCNOT
havebasis = "X"
, whereasControlledPhaseShift
andRZ
havebasis = "Z"
.- Type:
str or None
- batch_size¶
Batch size of the operator if it is used with broadcasted parameters.
The
batch_size
is determined based onndim_params
and the provided parameters for the operator. If (some of) the latter have an additional dimension, and this dimension has the same size for all parameters, its size is the batch size of the operator. If no parameter has an additional dimension, the batch size isNone
.- Returns:
Size of the parameter broadcasting dimension if present, else
None
.- Return type:
int or None
- control¶
The control wires.
- control_wires¶
Control wires of the operator.
For operations that are not controlled, this is an empty
Wires
object of length0
.- Returns:
The control wires of the operation.
- Return type:
- data¶
Create data property
- grad_method¶
Gradient computation method.
'A'
: analytic differentiation using the parameter-shift method.'F'
: finite difference numerical differentiation.None
: the operation may not be differentiated.
Default is
'F'
, orNone
if the Operation has zero parameters.
- grad_recipe = None¶
Gradient recipe for the parameter-shift method.
This is a tuple with one nested list per operation parameter. For parameter ϕk, the nested list contains elements of the form [ci,ai,si] where i is the index of the term, resulting in a gradient recipe of
∂∂ϕkf=∑icif(aiϕk+si).If
None
, the default gradient recipe containing the two terms [c0,a0,s0]=[1/2,1,π/2] and [c1,a1,s1]=[−1/2,1,−π/2] is assumed for every parameter.- Type:
tuple(Union(list[list[float]], None)) or None
- has_adjoint = False¶
- has_decomposition = True¶
- has_diagonalizing_gates = False¶
- has_generator = False¶
- has_matrix = False¶
- has_qfunc_decomposition = False¶
- has_sparse_matrix = False¶
- hash¶
Integer hash that uniquely represents the operator.
- Type:
int
- hyperparameters¶
Dictionary of non-trainable variables that this operation depends on.
- Type:
dict
- id¶
Custom string to label a specific operator instance.
- is_hermitian¶
This property determines if an operator is hermitian.
- name¶
String for the name of the operator.
- ndim_params¶
Number of dimensions per trainable parameter of the operator.
By default, this property returns the numbers of dimensions of the parameters used for the operator creation. If the parameter sizes for an operator subclass are fixed, this property can be overwritten to return the fixed value.
- Returns:
Number of dimensions for each trainable parameter.
- Return type:
tuple
- num_params¶
Number of trainable parameters that the operator depends on.
By default, this property returns as many parameters as were used for the operator creation. If the number of parameters for an operator subclass is fixed, this property can be overwritten to return the fixed value.
- Returns:
number of parameters
- Return type:
int
- num_wires = None¶
Number of wires the operator acts on.
- ops¶
Operations to be applied.
- parameter_frequencies¶
Returns the frequencies for each operator parameter with respect to an expectation value of the form ⟨ψ|U(p)†ˆOU(p)|ψ⟩.
These frequencies encode the behaviour of the operator U(p) on the value of the expectation value as the parameters are modified. For more details, please see the
pennylane.fourier
module.- Returns:
Tuple of frequencies for each parameter. Note that only non-negative frequency values are returned.
- Return type:
list[tuple[int or float]]
Example
>>> op = qml.CRot(0.4, 0.1, 0.3, wires=[0, 1]) >>> op.parameter_frequencies [(0.5, 1), (0.5, 1), (0.5, 1)]
For operators that define a generator, the parameter frequencies are directly related to the eigenvalues of the generator:
>>> op = qml.ControlledPhaseShift(0.1, wires=[0, 1]) >>> op.parameter_frequencies [(1,)] >>> gen = qml.generator(op, format="observable") >>> gen_eigvals = qml.eigvals(gen) >>> qml.gradients.eigvals_to_frequencies(tuple(gen_eigvals)) (1.0,)
For more details on this relationship, see
eigvals_to_frequencies()
.
- parameters¶
Trainable parameters that the operator depends on.
- pauli_rep¶
A
PauliSentence
representation of the Operator, orNone
if it doesn’t have one.
- resource_keys = {'ops'}¶
- resource_params¶
- target_wires¶
The wires of the target operators.
- wires¶
All wires involved in the operation.
Methods
adjoint
()Create an operation that is the adjoint of this one.
compute_decomposition
(ops, control)Representation of the operator as a product of other operators (static method).
compute_diagonalizing_gates
(*params, wires, ...)Sequence of gates that diagonalize the operator in the computational basis (static method).
compute_eigvals
(*params, **hyperparams)Eigenvalues of the operator in the computational basis (static method).
compute_matrix
(*params, **hyperparams)Representation of the operator as a canonical matrix in the computational basis (static method).
compute_qfunc_decomposition
(*args, ...)Experimental method to compute the dynamic decomposition of the operator with program capture enabled.
compute_sparse_matrix
(*params[, format])Representation of the operator as a sparse matrix in the computational basis (static method).
Representation of the operator as a product of other operators.
Sequence of gates that diagonalize the operator in the computational basis.
eigvals
()Eigenvalues of the operator in the computational basis.
Generator of an operator that is in single-parameter-form.
label
([decimals, base_label, cache])A customizable string representation of the operator.
map_wires
(wire_map)Returns a copy of the current operator with its wires changed according to the given wire map.
matrix
([wire_order])Representation of the operator as a matrix in the computational basis.
pow
(z)A list of new operators equal to this one raised to the given power.
queue
([context])Append the operator to the Operator queue.
simplify
()Reduce the depth of nested operators to the minimum.
The parameters required to implement a single-qubit gate as an equivalent
Rot
gate, up to a global phase.sparse_matrix
([wire_order, format])Representation of the operator as a sparse matrix in the computational basis.
terms
()Representation of the operator as a linear combination of other operators.
- adjoint()¶
Create an operation that is the adjoint of this one.
Adjointed operations are the conjugated and transposed version of the original operation. Adjointed ops are equivalent to the inverted operation for unitary gates.
- Returns:
The adjointed operation.
- static compute_decomposition(ops, control)[source]¶
Representation of the operator as a product of other operators (static method).
O=O1O2…On.Note
Operations making up the decomposition should be queued within the
compute_decomposition
method.See also
- Parameters:
ops (list[Operator]) – operations to apply
control (Sequence[int]) – the wires controlling which operation is applied
- Returns:
decomposition of the operator
- Return type:
list[Operator]
Example
>>> ops = [qml.X(2), qml.X(3), qml.Y(2), qml.SWAP([2,3])] >>> qml.Select.compute_decomposition(ops, control=[0,1]) [MultiControlledX(wires=[0, 1, 2], control_values=[0, 0]), MultiControlledX(wires=[0, 1, 3], control_values=[0, 1), Controlled(Y(2), control_wires=[0, 1], control_values=[True, False]), Controlled(SWAP(wires=[2, 3]), control_wires=[0, 1])]
- static compute_diagonalizing_gates(*params, wires, **hyperparams)¶
Sequence of gates that diagonalize the operator in the computational basis (static method).
Given the eigendecomposition O=UΣU† where Σ is a diagonal matrix containing the eigenvalues, the sequence of diagonalizing gates implements the unitary U†.
The diagonalizing gates rotate the state into the eigenbasis of the operator.
See also
- Parameters:
params (list) – trainable parameters of the operator, as stored in the
parameters
attributewires (Iterable[Any], Wires) – wires that the operator acts on
hyperparams (dict) – non-trainable hyperparameters of the operator, as stored in the
hyperparameters
attribute
- Returns:
list of diagonalizing gates
- Return type:
list[.Operator]
- static compute_eigvals(*params, **hyperparams)¶
Eigenvalues of the operator in the computational basis (static method).
If
diagonalizing_gates
are specified and implement a unitary U†, the operator can be reconstructed asO=UΣU†,where Σ is the diagonal matrix containing the eigenvalues.
Otherwise, no particular order for the eigenvalues is guaranteed.
See also
- Parameters:
*params (list) – trainable parameters of the operator, as stored in the
parameters
attribute**hyperparams (dict) – non-trainable hyperparameters of the operator, as stored in the
hyperparameters
attribute
- Returns:
eigenvalues
- Return type:
tensor_like
- static compute_matrix(*params, **hyperparams)¶
Representation of the operator as a canonical matrix in the computational basis (static method).
The canonical matrix is the textbook matrix representation that does not consider wires. Implicitly, this assumes that the wires of the operator correspond to the global wire order.
See also
- Parameters:
*params (list) – trainable parameters of the operator, as stored in the
parameters
attribute**hyperparams (dict) – non-trainable hyperparameters of the operator, as stored in the
hyperparameters
attribute
- Returns:
matrix representation
- Return type:
tensor_like
- static compute_qfunc_decomposition(*args, **hyperparameters)¶
Experimental method to compute the dynamic decomposition of the operator with program capture enabled.
When the program capture feature is enabled with
qml.capture.enable()
, the decomposition of the operator is computed with this method if it is defined. Otherwise, thecompute_decomposition()
method is used.The exception to this rule is when the operator is returned from the
compute_decomposition()
method of another operator, in which case the decomposition is performed withcompute_decomposition()
(even if this method is defined), and not with this method.When
compute_qfunc_decomposition
is defined for an operator, the control flow operations within the method (specifying the decomposition of the operator) are recorded in the JAX representation.Note
This method is experimental and subject to change.
See also
- Parameters:
*args (list) – positional arguments passed to the operator, including trainable parameters and wires
**hyperparameters (dict) – non-trainable hyperparameters of the operator, as stored in the
hyperparameters
attribute
- static compute_sparse_matrix(*params, format='csr', **hyperparams)¶
Representation of the operator as a sparse matrix in the computational basis (static method).
The canonical matrix is the textbook matrix representation that does not consider wires. Implicitly, this assumes that the wires of the operator correspond to the global wire order.
See also
- Parameters:
*params (list) – trainable parameters of the operator, as stored in the
parameters
attributeformat (str) – format of the returned scipy sparse matrix, for example ‘csr’
**hyperparams (dict) – non-trainable hyperparameters of the operator, as stored in the
hyperparameters
attribute
- Returns:
sparse matrix representation
- Return type:
scipy.sparse._csr.csr_matrix
- decomposition()[source]¶
Representation of the operator as a product of other operators.
O=O1O2…OnA
DecompositionUndefinedError
is raised if no representation by decomposition is defined.See also
- Returns:
decomposition of the operator
- Return type:
list[Operator]
Example
>>> ops = [qml.X(2), qml.X(3), qml.Y(2), qml.SWAP([2,3])] >>> op = qml.Select(ops, control=[0,1]) >>> op.decomposition() [MultiControlledX(wires=[0, 1, 2], control_values=[0, 0]), MultiControlledX(wires=[0, 1, 3], control_values=[0, 1]), Controlled(Y(2), control_wires=[0, 1], control_values=[True, False]), Controlled(SWAP(wires=[2, 3]), control_wires=[0, 1])]
- diagonalizing_gates()¶
Sequence of gates that diagonalize the operator in the computational basis.
Given the eigendecomposition O=UΣU† where Σ is a diagonal matrix containing the eigenvalues, the sequence of diagonalizing gates implements the unitary U†.
The diagonalizing gates rotate the state into the eigenbasis of the operator.
A
DiagGatesUndefinedError
is raised if no representation by decomposition is defined.See also
- Returns:
a list of operators
- Return type:
list[.Operator] or None
- eigvals()¶
Eigenvalues of the operator in the computational basis.
If
diagonalizing_gates
are specified and implement a unitary U†, the operator can be reconstructed asO=UΣU†,where Σ is the diagonal matrix containing the eigenvalues.
Otherwise, no particular order for the eigenvalues is guaranteed.
Note
When eigenvalues are not explicitly defined, they are computed automatically from the matrix representation. Currently, this computation is not differentiable.
A
EigvalsUndefinedError
is raised if the eigenvalues have not been defined and cannot be inferred from the matrix representation.See also
- Returns:
eigenvalues
- Return type:
tensor_like
- generator()¶
Generator of an operator that is in single-parameter-form.
For example, for operator
U(ϕ)=eiϕ(0.5Y+Z⊗X)we get the generator
>>> U.generator() 0.5 * Y(0) + Z(0) @ X(1)
The generator may also be provided in the form of a dense or sparse Hamiltonian (using
LinearCombination
andSparseHamiltonian
respectively).
- label(decimals=None, base_label=None, cache=None)¶
A customizable string representation of the operator.
- Parameters:
decimals=None (int) – If
None
, no parameters are included. Else, specifies how to round the parameters.base_label=None (str) – overwrite the non-parameter component of the label
cache=None (dict) – dictionary that carries information between label calls in the same drawing
- Returns:
label to use in drawings
- Return type:
str
Example:
>>> op = qml.RX(1.23456, wires=0) >>> op.label() "RX" >>> op.label(base_label="my_label") "my_label" >>> op = qml.RX(1.23456, wires=0, id="test_data") >>> op.label() "RX("test_data")" >>> op.label(decimals=2) "RX\n(1.23,"test_data")" >>> op.label(base_label="my_label") "my_label("test_data")" >>> op.label(decimals=2, base_label="my_label") "my_label\n(1.23,"test_data")"
If the operation has a matrix-valued parameter and a cache dictionary is provided, unique matrices will be cached in the
'matrices'
key list. The label will contain the index of the matrix in the'matrices'
list.>>> op2 = qml.QubitUnitary(np.eye(2), wires=0) >>> cache = {'matrices': []} >>> op2.label(cache=cache) 'U(M0)' >>> cache['matrices'] [tensor([[1., 0.], [0., 1.]], requires_grad=True)] >>> op3 = qml.QubitUnitary(np.eye(4), wires=(0,1)) >>> op3.label(cache=cache) 'U(M1)' >>> cache['matrices'] [tensor([[1., 0.], [0., 1.]], requires_grad=True), tensor([[1., 0., 0., 0.], [0., 1., 0., 0.], [0., 0., 1., 0.], [0., 0., 0., 1.]], requires_grad=True)]
- map_wires(wire_map)[source]¶
Returns a copy of the current operator with its wires changed according to the given wire map.
- Parameters:
wire_map (dict) – dictionary containing the old wires as keys and the new wires as values
- Returns:
new operator
- Return type:
.Operator
- matrix(wire_order=None)¶
Representation of the operator as a matrix in the computational basis.
If
wire_order
is provided, the numerical representation considers the position of the operator’s wires in the global wire order. Otherwise, the wire order defaults to the operator’s wires.If the matrix depends on trainable parameters, the result will be cast in the same autodifferentiation framework as the parameters.
A
MatrixUndefinedError
is raised if the matrix representation has not been defined.See also
- Parameters:
wire_order (Iterable) – global wire order, must contain all wire labels from the operator’s wires
- Returns:
matrix representation
- Return type:
tensor_like
- pow(z)¶
A list of new operators equal to this one raised to the given power.
- Parameters:
z (float) – exponent for the operator
- Returns:
list[
Operator
]
- queue(context=<class 'pennylane.queuing.QueuingManager'>)¶
Append the operator to the Operator queue.
- simplify()¶
Reduce the depth of nested operators to the minimum.
- Returns:
simplified operator
- Return type:
.Operator
- single_qubit_rot_angles()¶
The parameters required to implement a single-qubit gate as an equivalent
Rot
gate, up to a global phase.- Returns:
A list of values [ϕ,θ,ω] such that RZ(ω)RY(θ)RZ(ϕ) is equivalent to the original operation.
- Return type:
tuple[float, float, float]
- sparse_matrix(wire_order=None, format='csr')¶
Representation of the operator as a sparse matrix in the computational basis.
If
wire_order
is provided, the numerical representation considers the position of the operator’s wires in the global wire order. Otherwise, the wire order defaults to the operator’s wires.A
SparseMatrixUndefinedError
is raised if the sparse matrix representation has not been defined.See also
- Parameters:
wire_order (Iterable) – global wire order, must contain all wire labels from the operator’s wires
format (str) – format of the returned scipy sparse matrix, for example ‘csr’
- Returns:
sparse matrix representation
- Return type:
scipy.sparse._csr.csr_matrix
- terms()¶
Representation of the operator as a linear combination of other operators.
O=∑iciOiA
TermsUndefinedError
is raised if no representation by terms is defined.- Returns:
list of coefficients ci and list of operations Oi
- Return type:
tuple[list[tensor_like or float], list[.Operation]]