PennyLane is a cross-platform Python library for quantum computing, quantum machine learning, and quantum chemistry. Train a quantum computer the same way as a neural network.
APACHE-2.0 License
Bot releases are hidden (Show)
A new default.tensor
device is now available for performing tensor network and matrix product state simulations of quantum circuits using the quimb backend. (#5699) (#5744) (#5786) (#5795)
Either method can be selected when instantiating the default.tensor
device by setting the method
keyword argument to "tn"
(tensor network) or "mps"
(matrix product state).
There are several templates in PennyLane that are tensor-network focused, which are excellent candidates for the "tn"
method for default.tensor
. The following example shows how a circuit comprising gates in a tree tensor network architecture can be efficiently simulated using method="tn"
.
import pennylane as qml
n_wires = 16
dev = qml.device("default.tensor", method="tn")
def block(weights, wires):
qml.CNOT(wires=[wires[0], wires[1]])
qml.RY(weights[0], wires=wires[0])
qml.RY(weights[1], wires=wires[1])
n_block_wires = 2
n_params_block = 2
n_blocks = qml.TTN.get_n_blocks(range(n_wires), n_block_wires)
template_weights = [[0.1, -0.3]] * n_blocks
@qml.qnode(dev)
def circuit(template_weights):
for i in range(n_wires):
qml.Hadamard(i)
qml.TTN(range(n_wires), n_block_wires, block, n_params_block, template_weights)
return qml.expval(qml.Z(n_wires - 1))
>>> circuit(template_weights)
0.3839174759751649
For matrix product state simulations (method="mps"
), we can make the execution be approximate by setting max_bond_dim
(see the device's documentation for more details). The maximum bond dimension has implications for the speed of the simulation and lets us control the degree of the approximation, as shown in the following example. First, set up the circuit:
import numpy as np
n_layers = 10
n_wires = 10
initial_shape, weights_shape = qml.SimplifiedTwoDesign.shape(n_layers, n_wires)
np.random.seed(1967)
initial_layer_weights = np.random.random(initial_shape)
weights = np.random.random(weights_shape)
def f():
qml.SimplifiedTwoDesign(initial_layer_weights, weights, range(n_wires))
return qml.expval(qml.Z(0))
The default.tensor
device is instantiated with a max_bond_dim
value:
dev_dq = qml.device("default.qubit")
value_dq = qml.QNode(f, dev_dq)()
dev_mps = qml.device("default.tensor", max_bond_dim=5)
value_mps = qml.QNode(f, dev_mps)()
With this bond dimension, the expectation values calculated for default.qubit
and default.tensor
are different:
>>> np.abs(value_dq - value_mps)
tensor(0.0253213, requires_grad=True)
Learn more about default.tensor
and how to configure it by visiting the how-to guide.
Support for building noise models and applying them to a quantum circuit has been added via the NoiseModel
class and an add_noise
transform. (#5674) (#5684) (#5718)
Under the hood, PennyLane's approach to noise models is insertion-based, meaning that noise is included by inserting additional operators (gates or channels) that describe the noise into the quantum circuit. Creating a NoiseModel
boils down to defining Boolean conditions under which specific noisy operations are inserted. There are several ways to specify conditions for adding noisy operations:
qml.noise.op_eq(op)
: if the operator op
is encountered in the circuit, add noise.
qml.noise.op_in(ops)
: if any operators in ops
are encountered in the circuit, add noise.
qml.noise.wires_eq(wires)
: if an operator is applied to wires
, add noise.
qml.noise.wires_in(wires)
: if an operator is applied to any wire in wires
, add noise.
custom noise conditions: custom conditions can be defined as functions decorated with qml.BooleanFn
that return a Boolean value. For example, the following function will insert noise if a qml.RY
operator is encountered with an angle of rotation that is less than 0.5
:
@qml.BooleanFn
def c0(op):
return isinstance(op, qml.RY) and op.parameters[0] < 0.5
Conditions can also be combined together with &
, and
, |
, etc. Once the conditions under which noise is to be inserted have been stated, we can specify exactly what noise is inserted with the following:
qml.noise.partial_wires(op)
: insert op
on the wires that are specified by the condition that triggers adding this noisedef n0(op, **kwargs):
qml.RY(op.parameters[0] * 0.05, wires=op.wires)
With that, we can create a qml.NoiseModel
object whose argument must be a dictionary mapping conditions to noise:
c1 = qml.noise.op_eq(qml.X) & qml.noise.wires_in([0, 1])
n1 = qml.noise.partial_wires(qml.AmplitudeDamping, 0.4)
noise_model = qml.NoiseModel({c0: n0, c1: n1})
>>> noise_model
NoiseModel({ BooleanFn(c0): n0 OpEq(PauliX) | WiresIn([0, 1]): AmplitudeDamping(gamma=0.4) })
The noise model created can then be added to a QNode with qml.add_noise
:
dev = qml.device("lightning.qubit", wires=3)
@qml.qnode(dev)
def circuit():
qml.Y(0)
qml.CNOT([0, 1])
qml.RY(0.3, wires=2) # triggers c0
qml.X(1) # triggers c1
return qml.state()
>>> print(qml.draw(circuit)())
0: ──Y────────╭●────┤ State
1: ───────────╰X──X─┤ State
2: ──RY(0.30)───────┤ State
>>> circuit = qml.add_noise(circuit, noise_model)
>>> print(qml.draw(circuit)())
0: ──Y────────╭●───────────────────────────────────┤ State
1: ───────────╰X─────────X──AmplitudeDamping(0.40)─┤ State
2: ──RY(0.30)──RY(0.01)────────────────────────────┤ State
If more than one transform is applied to a QNode, control over when/where the add_noise
transform is applied in relation to the other transforms can be specified with the level
keyword argument. By default, add_noise
is applied after all the transforms that have been manually applied to the QNode until that point. To learn more about this new functionality, check out our noise module documentation and keep your eyes peeled for an in-depth demo!
The new PennyLane quantum debugger allows pausing simulation via the qml.breakpoint()
command and provides tools for analyzing quantum circuits during execution. (#5680) (#5749) (#5789)
This includes monitoring the circuit via measurements using qml.debug_state()
, qml.debug_probs()
, qml.debug_expval()
, and qml.debug_tape()
, stepping through the operations in a quantum circuit, and interactively adding operations during execution.
Including qml.breakpoint()
in a circuit will cause the simulation to pause during execution and bring up the interactive console. For example, consider the following code in a Python file called script.py
:
@qml.qnode(qml.device('default.qubit', wires=(0,1,2)))
def circuit(x):
qml.Hadamard(wires=0)
qml.CNOT(wires=(0,2))
qml.breakpoint()
qml.RX(x, wires=1)
qml.RY(x, wires=2)
qml.breakpoint()
return qml.sample()
circuit(1.2345)
Upon executing script.py
, the simulation pauses at the first breakpoint:
> /Users/your/path/to/script.py(8)circuit()
-> qml.RX(x, wires=1)
[pldb]
While debugging, we can access circuit information. For example, qml.debug_tape()
returns the tape of the circuit, giving access to its operations and drawing:
[pldb] tape = qml.debug_tape()
[pldb] print(tape.draw(wire_order=[0,1,2]))
0: ──H─╭●─┤
2: ────╰X─┤
[pldb] tape.operations
[Hadamard(wires=[0]), CNOT(wires=[0, 2])]
While qml.debug_state()
is equivalent to qml.state()
and gives the current state:
[pldb] print(qml.debug_state())
[0.70710678+0.j 0. +0.j 0. +0.j 0. +0.j 1. +0.j 0.70710678+0.j 0. +0.j 0. +0.j]
Other debugger functions like qml.debug_probs()
and qml.debug_expval()
also function like their simulation counterparts (qml.probs
and qml.expval
, respectively) and are described in more detail in the debugger documentation Additionally, standard debugging commands are available to navigate through code, including list
, longlist
, next
, continue
, and quit
, as described in the debugging documentation. Finally, to modify a circuit mid-run, simply call the desired PennyLane operations:
[pldb] qml.CNOT(wires=(0,2))
CNOT(wires=[0, 2])
[pldb] print(qml.debug_tape().draw(wire_order=[0,1,2]))
0: ──H─╭●─╭●─┤
2: ────╰X─╰X─┤
Stay tuned for an in-depth demonstration on using this feature with real-world examples! Convert between OpenFermion and PennyLane 🤝
Two new functions called qml.from_openfermion
and qml.to_openfermion
are now available to convert between OpenFermion and PennyLane objects. This includes both fermionic and qubit operators. (#5773) (#5808) (#5881)
For fermionic operators:
>>> import openfermion
>>> of_fermionic = openfermion.FermionOperator('0^ 2')
>>> type(of_fermionic)
<class 'openfermion.ops.operators.fermion_operator.FermionOperator'>
>>> pl_fermionic = qml.from_openfermion(of_fermionic)
>>> type(pl_fermionic)
<class 'pennylane.fermi.fermionic.FermiWord'>
>>> print(pl_fermionic)
a⁺(0) a(2)
And for qubit operators:
>>> of_qubit = 0.5 * openfermion.QubitOperator('X0 X5')
>>> pl_qubit = qml.from_openfermion(of_qubit)
>>> print(pl_qubit)
0.5 * (X(0) @ X(5))
It is now possible to control the stage at which qml.draw
, qml.draw_mpl
, and qml.specs
occur within a QNode's transform program. (#5855) (#5781)
Consider the following circuit which has multiple transforms applied:
@qml.transforms.split_non_commuting
@qml.transforms.cancel_inverses
@qml.transforms.merge_rotations
@qml.qnode(qml.device("default.qubit"))
def f():
qml.Hadamard(0)
qml.Y(0)
qml.RX(0.4, 0)
qml.RX(-0.4, 0)
qml.Y(0)
return qml.expval(qml.X(0) + 2 * qml.Y(0))
We can specify a level
value when using qml.draw()
:
>>> print(qml.draw(f, level=0)()) # input program
0: ──H──Y──RX(0.40)──RX(-0.40)──Y─┤ <X+(2.00*Y)>
>>> print(qml.draw(f, level=1)()) # rotations merged
0: ──H──Y──Y─┤ <X+(2.00*Y)>
>>> print(qml.draw(f, level=2)()) # inverses cancelled
0: ──H─┤ <X+(2.00*Y)>
>>> print(qml.draw(f, level=3)()) # Hamiltonian expanded
0: ──H─┤ <X>
0: ──H─┤ <Y>
The qml.workflow.get_transform_program function can be used to see the full transform program.
>>> qml.workflow.get_transform_program(f)
TransformProgram(merge_rotations, cancel_inverses, split_non_commuting, validate_device_wires, mid_circuit_measurements, decompose, validate_measurements, validate_observables, no_sampling)
Note that additional transforms can be added automatically from device preprocessing or gradient calculations. Rather than providing an integer value to level
, it is possible to target the "user"
, "gradient"
or "device"
stages:
n_wires = 3
x = np.random.random((2, n_wires))
@qml.qnode(qml.device("default.qubit"))
def f():
qml.BasicEntanglerLayers(x, range(n_wires))
return qml.expval(qml.X(0))
>>> print(qml.draw(f, level="device")())
0: ──RX(0.28)─╭●────╭X──RX(0.70)─╭●────╭X─┤ <X>
1: ──RX(0.52)─╰X─╭●─│───RX(0.65)─╰X─╭●─│──┤
2: ──RX(0.00)────╰X─╰●──RX(0.03)────╰X─╰●─┤
default.clifford
now supports arbitrary state-based measurements with qml.Snapshot
. (#5794)
qml.equal
now properly handles Pow
, Adjoint
, Exp
, and SProd
operators as arguments across different interfaces and tolerances with the addition of four new keyword arguments: check_interface
, check_trainability
, atol
and rtol
. (#5668)
The implementation for qml.assert_equal
has been updated for Operator
, Controlled
, Adjoint
, Pow
, Exp
, SProd
, ControlledSequence
, Prod
, Sum
, Tensor
and Hamiltonian
instances. (#5780) (#5877)
qml.from_qasm
now supports the ability to convert mid-circuit measurements from OpenQASM 2
code, and it can now also take an optional argument to specify a list of measurements to be performed at the end of the circuit, just like qml.from_qiskit
. (#5818)
Four new operators have been added for simulating noise on the default.qutrit.mixed
device: (#5502) (#5793) (#5503) (#5757) (#5799) (#5784)
qml.QutritDepolarizingChannel
: a channel that adds depolarizing noise.qml.QutritChannel
: enables the specification of noise using a collection of (3x3) Kraus matrices.qml.QutritAmplitudeDamping
: a channel that adds noise processes modelled by amplitude damping.qml.TritFlip
: a channel that adds trit flip errors, such as misclassification.The default.qubit
device supports a depth-first tree-traversal algorithm to accelerate native mid-circuit measurement execution. Accessible through the QNode argument mcm_method="tree-traversal"
, this new implementation supports classical control, collecting statistics, and post-selection, along with all measurements enabled with qml.dynamic_one_shot
. More information about this new mid-circuit measurement method can be found on our measurement documentation page. (#5180)
qml.QNode
and the @qml.qnode
decorator now accept two new keyword arguments: postselect_mode
and mcm_method
. These keyword arguments can be used to configure how the device should behave when running circuits with mid-circuit measurements. (#5679) (#5833) (#5850)
postselect_mode="hw-like"
indicates to devices to discard invalid shots when postselecting mid-circuit measurements. Use postselect_mode="fill-shots"
to unconditionally sample the postselected value, thus making all samples valid. This is equivalent to sampling until the number of valid samples matches the total number of shots.mcm_method
will indicate which strategy to use for running circuits with mid-circuit measurements. Use mcm_method="deferred"
to use the deferred measurements principle, or mcm_method="one-shot"
to execute once for each shot. If qml.qjit
is being used (the Catalyst compiler), mcm_method="single-branch-statistics"
is also available. Using this method, a single branch of the execution tree will be randomly explored.The dynamic_one_shot
transform received a few improvements:
When using defer_measurements
with postselection, operations that will never be active due to the postselected state are skipped in the transformed quantum circuit. In addition, postselected controls are skipped, as they are evaluated when the transform is applied. This optimization feature can be turned off by setting reduce_postselected=False
. (#5558)
Consider a simple circuit with three mid-circuit measurements, two of which are postselecting, and a single gate conditioned on those measurements:
@qml.qnode(qml.device("default.qubit"))
def node(x):
qml.RX(x, 0)
qml.RX(x, 1)
qml.RX(x, 2)
mcm0 = qml.measure(0, postselect=0, reset=False)
mcm1 = qml.measure(1, postselect=None, reset=True)
mcm2 = qml.measure(2, postselect=1, reset=False)
qml.cond(mcm0 + mcm1 + mcm2 == 1, qml.RX)(0.5, 3)
return qml.expval(qml.Z(0) @ qml.Z(3))
Without the new optimization, we obtain three gates, each controlled on the three measured qubits. They correspond to the combinations of controls that satisfy the condition mcm0 + mcm1 + mcm2 == 1
:
>>> print(qml.draw(qml.defer_measurements(node, reduce_postselected=False))(0.6))
0: ──RX(0.60)──|0⟩⟨0|─╭●─────────────────────────────────────────────┤ ╭<Z@Z>
1: ──RX(0.60)─────────│──╭●─╭X───────────────────────────────────────┤ │
2: ──RX(0.60)─────────│──│──│───|1⟩⟨1|─╭○────────╭○────────╭●────────┤ │
3: ───────────────────│──│──│──────────├RX(0.50)─├RX(0.50)─├RX(0.50)─┤ ╰<Z@Z>
4: ───────────────────╰X─│──│──────────├○────────├●────────├○────────┤
5: ──────────────────────╰X─╰●─────────╰●────────╰○────────╰○────────┤
If we do not explicitly deactivate the optimization, we obtain a much simpler circuit:
>>> print(qml.draw(qml.defer_measurements(node))(0.6))
0: ──RX(0.60)──|0⟩⟨0|─╭●─────────────────┤ ╭<Z@Z>
1: ──RX(0.60)─────────│──╭●─╭X───────────┤ │
2: ──RX(0.60)─────────│──│──│───|1⟩⟨1|───┤ │
3: ───────────────────│──│──│──╭RX(0.50)─┤ ╰<Z@Z>
4: ───────────────────╰X─│──│──│─────────┤
5: ──────────────────────╰X─╰●─╰○────────┤
There is only one controlled gate with only one control wire.
Mid-circuit measurement tests have been streamlined and refactored, removing most end-to-end tests from the native MCM test file, but keeping one that validates multiple mid-circuit measurements with any allowed return and interface end-to-end tests. (#5787)
The QROM algorithm is now available in PennyLane with qml.QROM
. This template allows you to enter classical data in the form of bitstrings. (#5688)
bitstrings = ["010", "111", "110", "000"]
dev = qml.device("default.qubit", shots = 1)
@qml.qnode(dev)
def circuit():
qml.BasisEmbedding(2, wires = [0,1])
qml.QROM(bitstrings = bitstrings, control_wires = [0,1], target_wires = [2,3,4], work_wires = [5,6,7])
return qml.sample(wires = [2,3,4])
>>> print(circuit())
[1 1 0]
A number of templates have been updated to be valid PyTrees and PennyLane operations. (#5698)
PennyLane operators, measurements, and QNodes can now automatically be captured as instructions in JAXPR. (#5564) (#5511) (#5708) (#5523) (#5686) (#5889)
The qml.PyTrees
module now has flatten
and unflatten
methods for serializing PyTrees. (#5701)
qml.sample
can now be used on Boolean values representing mid-circuit measurement results in traced quantum functions. This feature is used with Catalyst to enable the pattern m = measure(0); qml.sample(m)
. (#5673)
The qml.qchem.Molecule
object received a few improvements:
The qml.qchem.molecular_hamiltonian
function now supports parity and Bravyi-Kitaev mappings. (#5657)
qml.qchem.molecular_dipole
function has been added for calculating the dipole operator using the "dhf"
and "openfermion"
backends. (#5764)
The qchem module now has dedicated functions for calling the pyscf
and openfermion
backends and the molecular_hamiltonian
and molecular_dipole
functions have been moved to hamiltonian
and dipole
modules. (#5553) (#5863)
More fermionic-to-qubit tests have been added to cover cases when the mapped operator is different for various mapping schemes. (#5873)
Logging now allows for an easier opt-in across the stack and support has been extended to Catalyst. (#5528)
Three new Pytest markers have been added for easier management of our test suite: unit
, integration
and system
. (#5517)
qml.MultiControlledX
can now be decomposed even when no work_wires
are provided. The implementation returns $\mathcal{O}(\mbox{len(control wires)}^2)$ operations and is applicable for any multi-controlled unitary gate. This decomposition is provided in arXiv:quant-ph/9503016. (#5735)
A new function called expectation_value
has been added to qml.math
to calculate the expectation value of a matrix for pure states. (#4484)
>>> state_vector = [1/np.sqrt(2), 0, 1/np.sqrt(2), 0]
>>> operator_matrix = qml.matrix(qml.PauliZ(0), wire_order=[0,1])
>>> qml.math.expectation_value(operator_matrix, state_vector)
tensor(-2.23711432e-17+0.j, requires_grad=True)
param_shift
with the broadcast=True
option now supports shot vectors and multiple measurements. (#5667)
qml.TrotterProduct
is now compatible with resource tracking by inheriting from ResourcesOperation
. (#5680)
packaging
is now a required package in PennyLane. (#5769)
qml.ctrl
now works with tuple-valued control_values
when applied to any already controlled operation. (#5725)
The sorting order of parameter-shift terms is now guaranteed to resolve ties in the absolute value with the sign of the shifts. (#5582)
qml.transforms.split_non_commuting
can now handle circuits containing measurements of multi-term observables. (#5729) (#5838) (#5828) (#5869) (#5939) (#5945)
qml.devices.LegacyDevice
is now an alias for qml.Device
, so it is easier to distinguish it from qml.devices.Device
, which follows the new device API. (#5581)
The dtype
for eigvals
of X
, Y
, Z
and Hadamard
is changed from int
to float
, making them consistent with the other observables. The dtype
of the returned values when sampling these observables (e.g. qml.sample(X(0))
) is also changed to float
. (#5607)
The framework for the development of an assert_equal
function for testing operator comparison has been set up. (#5634) (#5858)
The decompose
transform has an error
keyword argument to specify the type of error that should be raised, allowing error types to be more consistent with the context the decompose
function is used in. (#5669)
Empty initialization of PauliVSpace
is permitted. (#5675)
qml.tape.QuantumScript
properties are only calculated when needed, instead of on initialization. This decreases the classical overhead by over 20%. Also, par_info
, obs_sharing_wires
, and obs_sharing_wires_id
are now public attributes. (#5696)
The qml.data
module now supports PyTree data types as dataset attributes (#5732)
qml.ops.Conditional
now inherits from qml.ops.SymbolicOp
, thus it inherits several useful common functionalities. Other properties such as adjoint and diagonalizing gates have been added using the base
properties. (##5772)
New dispatches for qml.ops.Conditional
and qml.MeasurementValue
have been added to qml.equal
. (##5772)
The qml.snapshots
transform now supports arbitrary devices by running a separate tape for each snapshot for unsupported devices. (#5805)
The qml.Snapshot
operator now accepts sample-based measurements for finite-shot devices. (#5805)
Device preprocess transforms now happen inside the ML boundary. (#5791)
Transforms applied to callables now use functools.wraps
to preserve the docstring and call signature of the original function. (#5857)
qml.qsvt()
now supports JAX arrays with angle conversions. (#5853)
The sorting order of parameter-shift terms is now guaranteed to resolve ties in the absolute value with the sign of the shifts. (#5583)
Passing shots
as a keyword argument to a QNode
initialization now raises an error instead of ignoring the input. (#5748)
A custom decomposition can no longer be provided to qml.QDrift
. Instead, apply the operations in your custom operation directly with qml.apply
. (#5698)
Sampling observables composed of X
, Y
, Z
and Hadamard
now returns values of type float
instead of int
. (#5607)
qml.is_commuting
no longer accepts the wire_map
argument, which does not bring any functionality. (#5660)
qml.from_qasm_file
has been removed. The user can open files and load their content using qml.from_qasm
. (#5659)
qml.load
has been removed in favour of more specific functions, such as qml.from_qiskit
, etc. (#5654)
qml.transforms.convert_to_numpy_parameters
is now a proper transform and its output signature has changed, returning a list of QuantumScript
s and a post-processing function instead of simply the transformed circuit. (#5693)
Controlled.wires
does not include self.work_wires
anymore. That can be accessed separately through Controlled.work_wires
. Consequently, Controlled.active_wires
has been removed in favour of the more common Controlled.wires
. (#5728)
The simplify
argument in qml.Hamiltonian
and qml.ops.LinearCombination
has been deprecated. Instead, qml.simplify()
can be called on the constructed operator. (#5677)
qml.transforms.map_batch_transform
has been deprecated, since a transform can be applied directly to a batch of tapes. (#5676)
The default behaviour of qml.from_qasm()
to remove measurements in the QASM code has been deprecated. Use measurements=[]
to keep this behaviour or measurements=None
to keep the measurements from the QASM code. (#5882) (#5904)
The qml.qchem
docs have been updated to showcase the new improvements. (#5758) (#5638)
Several links to other functions in measurement process docstrings have been fixed. (#5913)
Information about mid-circuit measurements has been moved from the measurements quickstart page to its own mid-circuit measurements quickstart page (#5870)
The documentation for the default.tensor
device has been added. (#5719)
A small typo was fixed in the docstring for qml.sample
. (#5685)
Typesetting for some of the documentation was fixed, (use of left/right delimiters, fractions, and fixing incorrectly set up commands) (#5804)
The qml.Tracker
examples have been updated. (#5803)
The input types for coupling_map
in qml.transpile
have been updated to reflect all the allowed input types by nx.to_networkx_graph
. (#5864)
The text in the qml.data
module and datasets quickstart has been slightly modified to lead to the quickstart first and highlight list_datasets
. (5484)
qml.compiler.active
first checks whether Catalyst is imported at all to avoid changing jax_enable_x64
on module initialization. (#5960)
The __invert__
dunder method of the MeasurementValue
class uses an array-valued function. (#5955)
Skip Projector
-measurement tests on devices that do not support it. (#5951)
The default.tensor
device now preserves the order of wires if the initial MPS is created from a dense state vector. (#5892)
Fixed a bug where hadamard_grad
returned a wrong shape for qml.probs()
without wires. (#5860)
An error is now raised on processing an AnnotatedQueue
into a QuantumScript
if the queue contains something other than an Operator
, MeasurementProcess
, or QuantumScript
. (#5866)
Fixed a bug in the wire handling on special controlled ops. (#5856)
Fixed a bug where Sum
's with repeated identical operations ended up with the same hash as Sum
's with different numbers of repeats. (#5851)
qml.qaoa.cost_layer
and qml.qaoa.mixer_layer
can now be used with Sum
operators. (#5846)
Fixed a bug where qml.MottonenStatePreparation
produces wrong derivatives at special parameter values. (#5774)
Fixed a bug where fractional powers and adjoints of operators were commuted, which is not well-defined/correct in general. Adjoints of fractional powers can no longer be evaluated. (#5835)
qml.qnn.TorchLayer
now works with tuple returns. (#5816)
An error is now raised if a transform is applied to a catalyst qjit object. (#5826)
qml.qnn.KerasLayer
and qml.qnn.TorchLayer
no longer mutate the input qml.QNode
's interface. (#5800)
Docker builds on PR merging has been disabled. (#5777)
The validation of the adjoint method in DefaultQubit
correctly handles device wires now. (#5761)
QuantumPhaseEstimation.map_wires
on longer modifies the original operation instance. (#5698)
The decomposition of qml.AmplitudeAmplification
now correctly queues all operations. (#5698)
Replaced semantic_version
with packaging.version.Version
, since the former cannot handle the metadata .post
in the version string. (#5754)
The dynamic_one_shot
transform now has expanded support for the jax
and torch
interfaces. (#5672)
The decomposition of StronglyEntanglingLayers
is now compatible with broadcasting. (#5716)
qml.cond
can now be applied to ControlledOp
operations when deferring measurements. (#5725)
The legacy Tensor
class can now handle a Projector
with abstract tracer input. (#5720)
Fixed a bug that raised an error regarding expected versus actual dtype
when using JAX-JIT
on a circuit that returned samples of observables containing the qml.Identity
operator. (#5607)
The signature of CaptureMeta
objects (like Operator
) now match the signature of the __init__
call. (#5727)
Vanilla NumPy arrays are now used in test_projector_expectation
to avoid differentiating qml.Projector
with respect to the state attribute. (#5683)
qml.Projector
is now compatible with jax.jit
. (#5595)
Finite-shot circuits with a qml.probs
measurement, both with a wires
or op
argument, can now be compiled with jax.jit
. (#5619)
param_shift
, finite_diff
, compile
, insert
, merge_rotations
, and transpile
now all work with circuits with non-commuting measurements. (#5424) (#5681)
A correction has been added to qml.bravyi_kitaev
to call the correct function for a qml.FermiSentence
input. (#5671)
Fixed a bug where sum_expand
produces incorrect result dimensions when combined with shot vectors, multiple measurements, and parameter broadcasting. (#5702)
Fixed a bug in qml.math.dot
that raises an error when only one of the operands is a scalar. (#5702)
qml.matrix
is now compatible with QNodes compiled by qml.qjit
. (#5753)
qml.snapshots
raises an error when a measurement other than qml.state
is requested from default.qubit.legacy
instead of silently returning the statevector. (#5805)
Fixed a bug where default.qutrit
was falsely determined to be natively compatible with qml.snapshots
. (#5805)
Fixed a bug where the measurement of a qml.Snapshot
instance was not passed on during the qml.adjoint
and qml.ctrl
operations. (#5805)
qml.CNOT
and qml.Toffoli
now have an arithmetic_depth
of 1
, as they are controlled operations. (#5797)
Fixed a bug where the gradient of ControlledSequence
, Reflection
, AmplitudeAmplification
, and Qubitization
was incorrect on default.qubit.legacy
with parameter_shift
. (#5806)
Fixed a bug where split_non_commuting
raises an error when the circuit contains measurements of observables that are not Pauli words. (#5827)
The simplify
method for qml.Exp
now returns an operator with the correct number of Trotter steps, i.e. equal to the one from the pre-simplified operator. (#5831)
Fixed a bug where CompositeOp.overlapping_ops
would put overlapping operators in different groups, leading to incorrect results returned by LinearCombination.eigvals()
. (#5847)
The correct decomposition for a qml.PauliRot
with an identity as pauli_word
has been implemented, i.e. returns a qml.GlobalPhase
with half the angle. (#5875)
qml.pauli_decompose
now works in a jit-ted context, such as jax.jit
and qml.qjit
. (#5878)
This release contains contributions from (in alphabetical order):
Tarun Kumar Allamsetty, Guillermo Alonso-Linaje, Utkarsh Azad, Lillian M. A. Frederiksen, Ludmila Botelho, Gabriel Bottrill, Thomas Bromley, Jack Brown, Astral Cai, Ahmed Darwish, Isaac De Vlugt, Diksha Dhawan, Pietropaolo Frisoni, Emiliano Godinez, Diego Guala, Daria Van Hende, Austin Huang, David Ittah, Soran Jahangiri, Rohan Jain, Mashhood Khan, Korbinian Kottmann, Christina Lee, Vincent Michaud-Rioux, Lee James O'Riordan, Mudit Pandey, Kenya Sakka, Jay Soni, Kazuki Tsuoka, Haochen Paul Wang, David Wierichs.
Published by astralcai 6 months ago
Two new user-facing classes enable calculating and propagating gate errors in PennyLane:
qml.resource.SpectralNormError
: the spectral norm error is defined as the distance, in spectral norm, between the true unitary we intend to apply and the approximate unitary that is actually applied.
qml.resource.ErrorOperation
: a base class that inherits from qml.operation.Operation
and represents quantum operations which carry some form of algorithmic error.
SpectralNormError
can be used for back-of-the-envelope type calculations like obtaining the spectral norm error between two unitaries via get_error
:
import pennylane as qml
from pennylane.resource import ErrorOperation, SpectralNormError
intended_op = qml.RY(0.40, 0)
actual_op = qml.RY(0.41, 0) # angle of rotation is slightly off
>>> SpectralNormError.get_error(intended_op, actual_op)
0.004999994791668309
SpectralNormError
is also a key tool to specify errors in larger quantum circuits:
For operations representing a major building block of an algorithm, we can create a custom operation that inherits from ErrorOperation
. This child class must override the error
method and should return a SpectralNormError
instance:
class MyErrorOperation(ErrorOperation):
def __init__(self, error_val, wires):
self.error_val = error_val
super().__init__(wires=wires)
def error(self):
return SpectralNormError(self.error_val)
In this toy example, MyErrorOperation
introduces an arbitrary SpectralNormError
when called in a QNode. It does not require a decomposition or matrix representation when used with null.qubit
(suggested for use with resource and error estimation since circuit executions are not required to calculate resources or errors).
dev = qml.device("null.qubit")
@qml.qnode(dev)
def circuit():
MyErrorOperation(0.1, wires=0)
MyErrorOperation(0.2, wires=1)
return qml.state()
The total spectral norm error of the circuit can be calculated using qml.specs
:
>>> qml.specs(circuit)()['errors']
{'SpectralNormError': SpectralNormError(0.30000000000000004)}
PennyLane already includes a number of built-in building blocks for algorithms like QuantumPhaseEstimation
and TrotterProduct
. TrotterProduct
now propagates errors based on the number of steps performed in the Trotter product. QuantumPhaseEstimation
now propagates errors based on the error of its input unitary.
dev = qml.device('null.qubit')
hamiltonian = qml.dot([1.0, 0.5, -0.25], [qml.X(0), qml.Y(0), qml.Z(0)])
@qml.qnode(dev)
def circuit():
qml.TrotterProduct(hamiltonian, time=0.1, order=2)
qml.QuantumPhaseEstimation(MyErrorOperation(0.01, wires=0), estimation_wires=[1, 2, 3])
return qml.state()
Again, the total spectral norm error of the circuit can be calculated using qml.specs
:
>>> qml.specs(circuit)()["errors"]
{'SpectralNormError': SpectralNormError(0.07616666666666666)}
Check out our error propagation demo to see how to use these new features in a real-world example!
The Fast Approximate BLock-Encodings (FABLE) algorithm for embedding a matrix into a quantum circuit as outlined in arXiv:2205.00081 is now accessible via the qml.FABLE
template. (#5107)
The usage of qml.FABLE
is similar to qml.BlockEncode
but provides a more efficient circuit construction at the cost of a user-defined approximation level, tol
. The number of wires that qml.FABLE
operates on is 2*n + 1
, where n
defines the dimension of the $2^n \times 2^n$ matrix that we want to block-encode.
import numpy as np
A = np.array([[0.1, 0.2], [0.3, 0.4]])
dev = qml.device('default.qubit', wires=3)
@qml.qnode(dev)
def circuit():
qml.FABLE(A, tol = 0.001, wires=range(3))
return qml.state()
>>> mat = qml.matrix(circuit)()
>>> 2 * mat[0:2, 0:2]
array([[0.1+0.j, 0.2+0.j], [0.3+0.j, 0.4+0.j]])
A high-level interface for amplitude amplification and its variants is now available via the new qml.AmplitudeAmplification
template. (#5160)
Based on arXiv:quant-ph/0005055, given a state $\vert \Psi \rangle = \alpha \vert \phi \rangle + \beta \vert \phi^{\perp} \rangle$, qml.AmplitudeAmplification
amplifies the amplitude of $\vert \phi \rangle$.
Here's an example with a target state $\vert \phi \rangle = \vert 2 \rangle = \vert 010 \rangle$, an input state $\vert \Psi \rangle = H^{\otimes 3} \vert 000 \rangle$, as well as an oracle that flips the sign of $\vert \phi \rangle$ and does nothing to $\vert \phi^{\perp} \rangle$, which can be achieved in this case through qml.FlipSign
.
@qml.prod
def generator(wires):
for wire in wires:
qml.Hadamard(wires=wire)
U = generator(wires=range(3))
O = qml.FlipSign(2, wires=range(3))
Here, U
is a quantum operation that is created by decorating a quantum function with @qml.prod
. This could alternatively be done by creating a user-defined custom operation with a decomposition. Amplitude amplification can then be set up within a circuit:
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circuit():
generator(wires=range(3)) # prepares |Psi> = U|0>
qml.AmplitudeAmplification(U, O, iters=10)
return qml.probs(wires=range(3))
>>> print(np.round(circuit(), 3))
[0.01 0.01 0.931 0.01 0.01 0.01 0.01 0.01 ]
As expected, we amplify the $\vert 2 \rangle$ state.
Reflecting about a given quantum state is now available via qml.Reflection
. This operation is very useful in the amplitude amplification algorithm and offers a generalization of qml.FlipSign
, which operates on basis states. (#5159)
qml.Reflection
works by providing an operation, $U$, that prepares the desired state, $\vert \psi \rangle$, that we want to reflect about. In other words, $U$ is such that $U \vert 0 \rangle = \vert \psi \rangle$. In PennyLane, $U$ must be an Operator
. For example, if we want to reflect about $\vert \psi \rangle = \vert + \rangle$, then $U = H$:
U = qml.Hadamard(wires=0)
dev = qml.device('default.qubit')
@qml.qnode(dev)
def circuit():
qml.Reflection(U)
return qml.state()
>>> circuit()
tensor([0.-6.123234e-17j, 1.+6.123234e-17j], requires_grad=True)
Performing qubitization is now easily accessible with the new qml.Qubitization
operator. (#5500)
qml.Qubitization
encodes a Hamiltonian into a suitable unitary operator. When applied in conjunction with quantum phase estimation (QPE), it allows for computing the eigenvalue of an eigenvector of the given Hamiltonian.
H = qml.dot([0.1, 0.3, -0.3], [qml.Z(0), qml.Z(1), qml.Z(0) @ qml.Z(2)])
@qml.qnode(qml.device("default.qubit"))
def circuit():
# initialize the eigenvector
qml.PauliX(2)
# apply QPE
measurements = qml.iterative_qpe(
qml.Qubitization(H, control = [3,4]), ancilla = 5, iters = 3
)
return qml.probs(op = measurements)
A new function called qml.bravyi_kitaev
has been added to perform the Bravyi-Kitaev mapping of fermionic Hamiltonians to qubit Hamiltonians. (#5390)
This function presents an alternative mapping to qml.jordan_wigner
or qml.parity_transform
which can help us measure expectation values more efficiently on hardware. Simply provide a fermionic Hamiltonian (created from from_string
, FermiA
, FermiC
, FermiSentence
, or FermiWord
) and the number of qubits / spin orbitals in the system, n
:
>>> fermi_ham = qml.fermi.from_string('0+ 1+ 1- 0-')
>>> qubit_ham = qml.bravyi_kitaev(fermi_ham, n=6, tol=0.0)
>>> print(qubit_ham)
0.25 * I(0) + -0.25 * Z(0) + -0.25 * (Z(0) @ Z(1)) + 0.25 * Z(1)
The qml.qchem.hf_state
function has been upgraded to be compatible with qml.parity_transform
and the new Bravyi-Kitaev mapping (qml.bravyi_kitaev
). (#5472) (#5472)
>>> state_bk = qml.qchem.hf_state(2, 6, basis="bravyi_kitaev")
>>> print(state_bk)
[1 0 0 0 0 0]
>>> state_parity = qml.qchem.hf_state(2, 6, basis="parity")
>>> print(state_parity)
[1 0 0 0 0 0]
The dynamical Lie algebra (DLA) of a set of operators captures the range of unitary evolutions that the operators can generate. In v0.36 of PennyLane, we have added support for calculating important DLA concepts including:
A new qml.lie_closure
function to compute the Lie closure of a list of operators, providing one way to obtain the DLA. (#5161) (#5169) (#5627)
For a list of operators ops = [op1, op2, op3, ..]
, one computes all nested commutators between ops
until no new operators are generated from commutation. All these operators together form the DLA, see e.g. section IIB of arXiv:2308.01432.
Take for example the following operators:
from pennylane import X, Y, Z
ops = [X(0) @ X(1), Z(0), Z(1)]
A first round of commutators between all elements yields the new operators Y(0) @ X(1)
and X(0) @ Y(1)
(omitting scalar prefactors).
>>> qml.commutator(X(0) @ X(1), Z(0))
-2j * (Y(0) @ X(1))
>>> qml.commutator(X(0) @ X(1), Z(1))
-2j * (X(0) @ Y(1))
A next round of commutators between all elements further yields the new operator Y(0) @ Y(1)
.
>>> qml.commutator(X(0) @ Y(1), Z(0))
-2j * (Y(0) @ Y(1))
After that, no new operators emerge from taking nested commutators and we have the resulting DLA. This can now be done in short via qml.lie_closure
as follows.
>>> ops = [X(0) @ X(1), Z(0), Z(1)]
>>> dla = qml.lie_closure(ops)
>>> dla
[X(0) @ X(1), Z(0), Z(1), -1.0 * (Y(0) @ X(1)), -1.0 * (X(0) @ Y(1)), -1.0 * (Y(0) @ Y(1))]
Computing the structure constants (the adjoint representation) of a dynamical Lie algebra. (5406)
For example, we can compute the adjoint representation of the transverse field Ising model DLA.
>>> dla = [X(0) @ X(1), Z(0), Z(1), Y(0) @ X(1), X(0) @ Y(1), Y(0) @ Y(1)]
>>> structure_const = qml.structure_constants(dla)
>>> structure_const.shape
(6, 6, 6)
Visit the documentation of qml.structure_constants to understand how structure constants are a useful way to represent a DLA.
Computing the center of a dynamical Lie algebra. (#5477)
Given a DLA g
, we can now compute its centre. The center
is the collection of operators that commute with all other operators in the DLA.
>>> g = [X(0), X(1) @ X(0), Y(1), Z(1) @ X(0)]
>>> qml.center(g)
[X(0)]
To help explain these concepts, check out the dynamical Lie algebras demo.
Mixed qutrit states can now be simulated with the default.qutrit.mixed
device. (#5495) (#5451) (#5186) (#5082) (#5213)
Thanks to contributors from the University of British Columbia, a mixed-state qutrit device is now available for simulation, providing a noise-capable equivalent to default.qutrit
.
dev = qml.device("default.qutrit.mixed")
def circuit():
qml.TRY(0.1, wires=0)
@qml.qnode(dev)
def shots_circuit():
circuit()
return qml.sample(), qml.expval(qml.GellMann(wires=0, index=1))
@qml.qnode(dev)
def density_matrix_circuit():
circuit()
return qml.state()
>>> shots_circuit(shots=5)
(array([0, 0, 0, 0, 0]), 0.19999999999999996)
>>> density_matrix_circuit()
tensor([[0.99750208+0.j, 0.04991671+0.j, 0. +0.j], [0.04991671+0.j, 0.00249792+0.j, 0. +0.j], [0. +0.j, 0. +0.j, 0. +0.j]], requires_grad=True)
However, there's one crucial ingredient that we still need to add: support for qutrit noise operations. Keep your eyes peeled for this to arrive in the coming releases!
This release completes the main phase of PennyLane's switchover to an updated approach for handling arithmetic operations between operators. The new approach is now enabled by default and is intended to realize a few objectives: 1. To make it as easy to work with PennyLane operators as it would be with pen and paper. 2. To improve the efficiency of operator arithmetic.
In many cases, this update should not break code. If issues do arise, check out the updated operator troubleshooting page and don't hesitate to reach out to us on the PennyLane discussion forum. As a last resort the old behaviour can be enabled by calling qml.operation.disable_new_opmath()
, but this is not recommended because support will not continue in future PennyLane versions (v0.36 and higher). (#5269)
A new class called qml.ops.LinearCombination
has been introduced. In essence, this class is an updated equivalent of the now-deprecated qml.ops.Hamiltonian
but for usage with the new operator arithmetic. (#5216)
qml.ops.Sum
now supports storing grouping information. Grouping type and method can be specified during construction using the grouping_type
and method
keyword arguments of qml.dot
, qml.sum
, or qml.ops.Sum
. The grouping indices are stored in Sum.grouping_indices
. (#5179)
a = qml.X(0)
b = qml.prod(qml.X(0), qml.X(1))
c = qml.Z(0)
obs = [a, b, c]
coeffs = [1.0, 2.0, 3.0]
op = qml.dot(coeffs, obs, grouping_type="qwc")
>>> op.grouping_indices
((2,), (0, 1))
Additionally, grouping_type
and method
can be set or changed after construction using Sum.compute_grouping()
:
a = qml.X(0)
b = qml.prod(qml.X(0), qml.X(1))
c = qml.Z(0)
obs = [a, b, c]
coeffs = [1.0, 2.0, 3.0]
op = qml.dot(coeffs, obs)
>>> op.grouping_indices is None
True
>>> op.compute_grouping(grouping_type="qwc")
>>> op.grouping_indices
((2,), (0, 1))
Note that the grouping indices refer to the lists returned by Sum.terms()
, not Sum.operands
.
A new function called qml.operation.convert_to_legacy_H
that converts Sum
, SProd
, and Prod
to Hamiltonian
instances has been added. This function is intended for developers and will be removed in a future release without a deprecation cycle. (#5309)
The qml.is_commuting
function now accepts Sum
, SProd
, and Prod
instances. (#5351)
Operators can now be left-multiplied by NumPy arrays (i.e., arr * op
). (#5361)
op.generator()
, where op
is an Operator
instance, now returns operators consistent with the global setting for qml.operator.active_new_opmath()
wherever possible. Sum
, SProd
and Prod
instances will be returned even after disabling the new operator arithmetic in cases where they offer additional functionality not available using legacy operators. (#5253) (#5410) (#5411) (#5421)
Prod
instances temporarily have a new obs
property, which helps smoothen the transition of the new operator arithmetic system. In particular, this is aimed at preventing breaking code that uses Tensor.obs
. The property has been immediately deprecated. Moving forward, we recommend using op.operands
. (#5539)
qml.ApproxTimeEvolution
is now compatible with any operator that has a defined pauli_rep
. (#5362)
Hamiltonian.pauli_rep
is now defined if the Hamiltonian is a linear combination of Pauli operators. (#5377)
Prod
instances created with qutrit operators now have a defined eigvals()
method. (#5400)
qml.transforms.hamiltonian_expand
and qml.transforms.sum_expand
can now handle multi-term observables with a constant offset (i.e., terms like qml.I()
). (#5414) (#5543)
qml.qchem.taper_operation
is now compatible with the new operator arithmetic. (#5326)
The warning for an observable that might not be hermitian in QNode executions has been removed. This enables jit-compilation. (#5506)
qml.transforms.split_non_commuting
will now work with single-term operator arithmetic. (#5314)
LinearCombination
and Sum
now accept _grouping_indices
on initialization. This addition is relevant to developers only. (#5524)
Calculating the dense, differentiable matrix for PauliSentence
and operators with Pauli sentences is now faster. (#5578)
ExpectationMP
, VarianceMP
, CountsMP
, and SampleMP
now have a process_counts
method (similar to process_samples
). This allows for calculating measurements given a counts
dictionary. (#5256) (#5395)
Type-hinting has been added in the Operator
class for better interpretability. (#5490)
An alternate strategy for sampling with multiple different shots
values has been implemented via the shots.bins()
method, which samples all shots at once and then processes each separately. (#5476)
A new module called qml.capture
that will contain PennyLane's own capturing mechanism for hybrid quantum-classical programs has been added. (#5509)
The dynamic_one_shot
transform has been introduced, enabling dynamic circuit execution on circuits with finite shots
and devices that natively support mid-circuit measurements. (#5266)
The QubitDevice
class and children classes support the dynamic_one_shot
transform provided that they support mid-circuit measurement operations natively. (#5317)
default.qubit
can now be provided a random seed for sampling mid-circuit measurements with finite shots. This (1) ensures that random behaviour is more consistent with dynamic_one_shot
and defer_measurements
and (2) makes our continuous-integration (CI) have less failures due to stochasticity. (#5337)
Gradient transforms may now be applied to batched/broadcasted QNodes as long as the broadcasting is in non-trainable parameters. (#5452)
The performance of computing the matrix of qml.QFT
has been improved. (#5351)
qml.transforms.broadcast_expand
now supports shot vectors when returning qml.sample()
. (#5473)
LightningVJPs
is now compatible with Lightning devices using the new device API. (#5469)
Obtaining classical shadows using the default.clifford
device is now compatible with stim v1.13.0
. (#5409)
default.mixed
has improved support for sampling-based measurements with non-NumPy interfaces. (#5514) (#5530)
default.mixed
now supports arbitrary state-based measurements with qml.Snapshot
. (#5552)
null.qubit
has been upgraded to the new device API and has support for all measurements and various modes of differentiation. (#5211)
Entanglement entropy can now be calculated with qml.math.vn_entanglement_entropy
, which computes the von Neumann entanglement entropy from a density matrix. A corresponding QNode transform, qml.qinfo.vn_entanglement_entropy
, has also been added. (#5306)
qml.draw
and qml.draw_mpl
will now attempt to sort the wires if no wire order is provided by the user or the device. (#5576)
A clear error message is added in KerasLayer
when using the newest version of TensorFlow with Keras 3 (which is not currently compatible with KerasLayer
), linking to instructions to enable Keras 2. (#5488)
qml.ops.Conditional
now stores the data
, num_params
, and ndim_param
attributes of the operator it wraps. (#5473)
The molecular_hamiltonian
function calls PySCF
directly when method='pyscf'
is selected. (#5118)
cache_execute
has been replaced with an alternate implementation based on @transform
. (#5318)
QNodes now defer diff_method
validation to the device under the new device API. (#5176)
The device test suite has been extended to cover gradient methods, templates and arithmetic observables. (#5273) (#5518)
A typo and string formatting mistake have been fixed in the error message for ClassicalShadow._convert_to_pauli_words
when the input is not a valid pauli_rep
. (#5572)
Circuits running on lightning.qubit
and that return qml.state()
now preserve the dtype
when specified. (#5547)
qml.matrix()
called on the following will now raise an error if wire_order
is not specified: * tapes with more than one wire * quantum functions * Operator
classes where num_wires
does not equal to 1 * QNodes if the device does not have wires specified. * PauliWord
s and PauliSentence
s with more than one wire. (#5328) (#5359)
single_tape_transform
, batch_transform
, qfunc_transform
, op_transform
, gradient_transform
and hessian_transform
have been removed. Instead, switch to using the new qml.transform
function. Please refer to the transform docs <https://docs.pennylane.ai/en/stable/code/qml_transforms.html#custom-transforms>
_ to see how this can be done. (#5339)
Attempting to multiply PauliWord
and PauliSentence
with *
will raise an error. Instead, use @
to conform with the PennyLane convention. (#5341)
DefaultQubit
now uses a pre-emptive key-splitting strategy to avoid reusing JAX PRNG keys throughout a single execute
call. (#5515)
qml.pauli.pauli_mult
and qml.pauli.pauli_mult_with_phase
have been removed. Instead, use qml.simplify(qml.prod(pauli_1, pauli_2))
to get the reduced operator. (#5324)
>>> op = qml.simplify(qml.prod(qml.PauliX(0), qml.PauliZ(0)))
>>> op -1j*(PauliY(wires=[0]))
>>> [phase], [base] = op.terms()
>>> phase, base
(-1j, PauliY(wires=[0]))
The dynamic_one_shot
transform now uses sampling (SampleMP
) to get back the values of the mid-circuit measurements. (#5486)
Operator
dunder methods now combine like-operator arithmetic classes via lazy=False
. This reduces the chances of getting a RecursionError
and makes nested operators easier to work with. (#5478)
The private functions _pauli_mult
, _binary_matrix
and _get_pauli_map
from the pauli
module have been removed. The same functionality can be achieved using newer features in the pauli
module. (#5323)
MeasurementProcess.name
and MeasurementProcess.data
have been removed. Use MeasurementProcess.obs.name
and MeasurementProcess.obs.data
instead. (#5321)
Operator.validate_subspace(subspace)
has been removed. Instead, use qml.ops.qutrit.validate_subspace(subspace)
. (#5311)
The contents of qml.interfaces
has been moved inside qml.workflow
. The old import path no longer exists. (#5329)
Since default.mixed
does not support snapshots with measurements, attempting to do so will result in a DeviceError
instead of getting the density matrix. (#5416)
LinearCombination._obs_data
has been removed. You can still use LinearCombination.compare
to check mathematical equivalence between a LinearCombination
and another operator. (#5504)
Accessing qml.ops.Hamiltonian
is deprecated because it points to the old version of the class that may not be compatible with the new approach to operator arithmetic. Instead, using qml.Hamiltonian
is recommended because it dispatches to the LinearCombination
class when the new approach to operator arithmetic is enabled. This will allow you to continue to use qml.Hamiltonian
with existing code without needing to make any changes. (#5393)
qml.load
has been deprecated. Instead, please use the functions outlined in the Importing workflows quickstart guide. (#5312)
Specifying control_values
with a bit string in qml.MultiControlledX
has been deprecated. Instead, use a list of booleans or 1s and 0s. (#5352)
qml.from_qasm_file
has been deprecated. Instead, please open the file and then load its content using qml.from_qasm
. (#5331)
>>> with open("test.qasm", "r") as f:
... circuit = qml.from_qasm(f.read())
A new page explaining the shapes and nesting of return types has been added. (#5418)
Redundant documentation for the evolve
function has been removed. (#5347)
The final example in the compile
docstring has been updated to use transforms correctly. (#5348)
A link to the demos for using qml.SpecialUnitary
and qml.QNGOptimizer
has been added to their respective docstrings. (#5376)
A code example in the qml.measure
docstring has been added that showcases returning mid-circuit measurement statistics from QNodes. (#5441)
The computational basis convention used for qml.measure
— 0 and 1 rather than ±1 — has been clarified in its docstring. (#5474)
A new Release news section has been added to the table of contents, containing release notes, deprecations, and other pages focusing on recent changes. (#5548)
A summary of all changes has been added in the "Updated Operators" page in the new "Release news" section in the docs. (#5483) (#5636)
Patches the QNode so that parameter-shift will be considered best with lightning if qml.metric_tensor
is in the transform program. (#5624)
Stopped printing the ID of qcut.MeasureNode
and qcut.PrepareNode
in tape drawing. (#5613)
Improves the error message for setting shots on the new device interface, or trying to access a property that no longer exists. (#5616)
Fixed a bug where qml.draw
and qml.draw_mpl
incorrectly raised errors for circuits collecting statistics on mid-circuit measurements while using qml.defer_measurements
. (#5610)
Using shot vectors with param_shift(... broadcast=True)
caused a bug. This combination is no longer supported and will be added again in the next release. Fixed a bug with custom gradient recipes that only consist of unshifted terms. (#5612) (#5623)
qml.counts
now returns the same keys with dynamic_one_shot
and defer_measurements
. (#5587)
null.qubit
now automatically supports any operation without a decomposition. (#5582)
Fixed a bug where the shape and type of derivatives obtained by applying a gradient transform to a QNode differed based on whether the QNode uses classical coprocessing. (#4945)
ApproxTimeEvolution
, CommutingEvolution
, QDrift
, and TrotterProduct
now de-queue their input observable. (#5524)
(In)equality of qml.HilbertSchmidt
instances is now reported correctly by qml.equal
. (#5538)
qml.ParticleConservingU1
and qml.ParticleConservingU2
no longer raise an error when the initial state is not specified but default to the all-zeros state. (#5535)
qml.counts
no longer returns negative samples when measuring 8 or more wires. (#5544) (#5556)
The dynamic_one_shot
transform now works with broadcasting. (#5473)
Diagonalizing gates are now applied when measuring qml.probs
on non-computational basis states on a Lightning device. (#5529)
two_qubit_decomposition
no longer diverges at a special case of a unitary matrix. (#5448)
The qml.QNSPSAOptimizer
now correctly handles optimization for legacy devices that do not follow the new device API. (#5497)
Operators applied to all wires are now drawn correctly in a circuit with mid-circuit measurements. (#5501)
Fixed a bug where certain unary mid-circuit measurement expressions would raise an uncaught error. (#5480)
Probabilities now sum to 1 when using the torch
interface with default_dtype
set to torch.float32
. (#5462)
Tensorflow can now handle devices with float32
results but float64
input parameters. (#5446)
Fixed a bug where the argnum
keyword argument of qml.gradients.stoch_pulse_grad
references the wrong parameters in a tape, creating an inconsistency with other differentiation methods and preventing some use cases. (#5458)
Bounded value failures due to numerical noise with calls to np.random.binomial
is now avoided. (#5447)
Using @
with legacy Hamiltonian instances now properly de-queues the previously existing operations. (#5455)
The QNSPSAOptimizer
now properly handles differentiable parameters, resulting in being able to use it for more than one optimization step. (#5439)
The QNode interface now resets if an error occurs during execution. (#5449)
Failing tests due to changes with Lightning's adjoint diff pipeline have been fixed. (#5450)
Failures occurring when making autoray-dispatched calls to Torch with paired CPU data have been fixed. (#5438)
jax.jit
now works with qml.sample
with a multi-wire observable. (#5422)
qml.qinfo.quantum_fisher
now works with non-default.qubit
devices. (#5423)
We no longer perform unwanted dtype
promotion in the pauli_rep
of SProd
instances when using Tensorflow. (#5246)
Fixed TestQubitIntegration.test_counts
in tests/interfaces/test_jax_qnode.py
to always produce counts for all outcomes. (#5336)
Fixed PauliSentence.to_mat(wire_order)
to support identities with wires. (#5407)
CompositeOp.map_wires
now correctly maps the overlapping_ops
property. (#5430)
DefaultQubit.supports_derivatives
has been updated to correctly handle circuits containing mid-circuit measurements and adjoint differentiation. (#5434)
SampleMP
, ExpectationMP
, CountsMP
, and VarianceMP
constructed with eigvals
can now properly process samples. (#5463)
Fixed a bug in hamiltonian_expand
that produces incorrect output dimensions when shot vectors are combined with parameter broadcasting. (#5494)
default.qubit
now allows measuring Identity
on no wires and observables containing Identity
on no wires. (#5570)
Fixed a bug where TorchLayer
does not work with shot vectors. (#5492)
Fixed a bug where the output shape of a QNode returning a list containing a single measurement is incorrect when combined with shot vectors. (#5492)
Fixed a bug in qml.math.kron
that makes Torch incompatible with NumPy. (#5540)
Fixed a bug in _group_measurements
that fails to group measurements with commuting observables when they are operands of Prod
. (#5525)
qml.equal
can now be used with sums and products that contain operators on no wires like I
and GlobalPhase
. (#5562)
CompositeOp.has_diagonalizing_gates
now does a more complete check of the base operators to ensure consistency between op.has_diagonalzing_gates
and op.diagonalizing_gates()
(#5603)
Updated the method
kwarg of qml.TrotterProduct().error()
to be more clear that we are computing upper-bounds. (#5637)
This release contains contributions from (in alphabetical order):
Tarun Kumar Allamsetty, Guillermo Alonso, Mikhail Andrenkov, Utkarsh Azad, Gabriel Bottrill, Thomas Bromley, Astral Cai, Diksha Dhawan, Isaac De Vlugt, Amintor Dusko, Pietropaolo Frisoni, Lillian M. A. Frederiksen, Diego Guala, Austin Huang, Soran Jahangiri, Korbinian Kottmann, Christina Lee, Vincent Michaud-Rioux, Mudit Pandey, Kenya Sakka, Jay Soni, Matthew Silverman, David Wierichs.
Published by Mandrenkov 7 months ago
Published by lillian542 8 months ago
This version of PennyLane makes it easier to import circuits from Qiskit. (#5218) (#5168)
The qml.from_qiskit
function converts a Qiskit QuantumCircuit into a PennyLane quantum function. Although qml.from_qiskit
already exists in PennyLane, we have made a number of improvements to make importing from Qiskit easier. And yes — qml.from_qiskit
functionality is compatible with both Qiskit 1.0 and earlier versions! Here's a comprehensive list of the improvements:
You can now append PennyLane measurements onto the quantum function returned by qml.from_qiskit
. Consider this simple Qiskit circuit:
import pennylane as qml
from qiskit import QuantumCircuit
qc = QuantumCircuit(2)
qc.rx(0.785, 0)
qc.ry(1.57, 1)
We can convert it into a PennyLane QNode in just a few lines, with PennyLane measurements
easily included:
>>> dev = qml.device("default.qubit")
>>> measurements = qml.expval(qml.Z(0) @ qml.Z(1))
>>> qfunc = qml.from_qiskit(qc, measurements=measurements)
>>> qnode = qml.QNode(qfunc, dev)
>>> qnode()
tensor(0.00056331, requires_grad=True)
Quantum circuits that already contain Qiskit-side measurements can be faithfully converted with qml.from_qiskit
. Consider this example Qiskit circuit:
qc = QuantumCircuit(3, 2) # Teleportation
qc.rx(0.9, 0) # Prepare input state on qubit 0
qc.h(1) # Prepare Bell state on qubits 1 and 2 qc.cx(1, 2)
qc.cx(0, 1) # Perform teleportation
qc.h(0)
qc.measure(0, 0)
qc.measure(1, 1)
with qc.if_test((1, 1)): # Perform first conditional
qc.x(2)
This circuit can be converted into PennyLane with the Qiskit measurements still accessible. For example, we can use those results as inputs to a mid-circuit measurement in PennyLane:
@qml.qnode(dev)
def teleport():
m0, m1 = qml.from_qiskit(qc)()
qml.cond(m0, qml.Z)(2)
return qml.density_matrix(2)
>>> teleport()
tensor([[0.81080498+0.j , 0. +0.39166345j],
[0. -0.39166345j, 0.18919502+0.j ]], requires_grad=True)
It is now more intuitive to handle and differentiate parametrized Qiskit circuits. Consider the following circuit:
from qiskit.circuit import Parameter
from pennylane import numpy as np
angle0 = Parameter("x") angle1 = Parameter("y")
qc = QuantumCircuit(2, 2)
qc.rx(angle0, 0)
qc.ry(angle1, 1)
qc.cx(1, 0)
We can convert this circuit into a QNode with two arguments, corresponding to x
and y
:
measurements = qml.expval(qml.PauliZ(0))
qfunc = qml.from_qiskit(qc, measurements)
qnode = qml.QNode(qfunc, dev)
The QNode can be evaluated and differentiated:
>>> x, y = np.array([0.4, 0.5], requires_grad=True)
>>> qnode(x, y)
tensor(0.80830707, requires_grad=True)
>>> qml.grad(qnode)(x, y)
(tensor(-0.34174675, requires_grad=True), tensor(-0.44158016, requires_grad=True))
This shows how easy it is to make a Qiskit circuit differentiable with PennyLane.
In addition to circuits, it is also possible to convert operators from Qiskit to PennyLane with a new function called qml.from_qiskit_op
. (#5251)
A Qiskit SparsePauliOp can be converted to a PennyLane operator using qml.from_qiskit_op
:
>>> from qiskit.quantum_info import SparsePauliOp
>>> qiskit_op = SparsePauliOp(["II", "XY"])
>>> qiskit_op SparsePauliOp(['II', 'XY'], coeffs=[1.+0.j, 1.+0.j])
>>> pl_op = qml.from_qiskit_op(qiskit_op)
>>> pl_op I(0) + X(1) @ Y(0)
Combined with qml.from_qiskit
, it becomes easy to quickly calculate quantities like expectation values by converting the whole workflow to PennyLane:
qc = QuantumCircuit(2) # Create circuit
qc.rx(0.785, 0)
qc.ry(1.57, 1)
measurements = qml.expval(pl_op) # Create QNode
qfunc = qml.from_qiskit(qc, measurements)
qnode = qml.QNode(qfunc, dev)
>>> qnode() # Evaluate!
tensor(0.29317504, requires_grad=True)
Mid-circuit measurements can now be more scalable and efficient in finite-shots mode with default.qubit
by simulating them in a similar way to what happens on quantum hardware. (#5088) (#5120)
Previously, mid-circuit measurements (MCMs) would be automatically replaced with an additional qubit using the @qml.defer_measurements
transform. The circuit below would have required thousands of qubits to simulate.
Now, MCMs are performed in a similar way to quantum hardware with finite shots on default.qubit
. For each shot and each time an MCM is encountered, the device evaluates the probability of projecting onto |0>
or |1>
and makes a random choice to collapse the circuit state. This approach works well when there are a lot of MCMs and the number of shots is not too high.
import pennylane as qml
dev = qml.device("default.qubit", shots=10)
@qml.qnode(dev)
def f():
for i in range(1967):
qml.Hadamard(0)
qml.measure(0)
return qml.sample(qml.PauliX(0))
>>> f()
tensor([-1, -1, -1, 1, 1, -1, 1, -1, 1, -1], requires_grad=True)
Over the past few releases, PennyLane's approach to operator arithmetic has been in the process of being overhauled. We have a few objectives:
The updated operator arithmetic functionality is still being finalized, but can be activated using qml.operation.enable_new_opmath()
. In the next release, the new behaviour will become the default, so we recommend enabling now to become familiar with the new system!
The following updates have been made in this version of PennyLane:
You can now easily access Pauli operators via I
, X
, Y
, and Z
: (#5116)
>>> from pennylane import I, X, Y, Z
>>> X(0) X(0) ```
The original long-form names `Identity`, `PauliX`, `PauliY`, and `PauliZ` remain available, but use of the short-form names is now recommended.
A new qml.commutator
function is now available that allows you to compute commutators between PennyLane operators. (#5051) (#5052) (#5098)
>>> qml.commutator(X(0), Y(0))
2j * Z(0)
Operators in PennyLane can have a backend Pauli representation, which can be used to perform faster operator arithmetic. Now, the Pauli representation will be automatically used for calculations when available. (#4989) (#5001) (#5003) (#5017) (#5027)
The Pauli representation can be optionally accessed via op.pauli_rep
:
>>> qml.operation.enable_new_opmath()
>>> op = X(0) + Y(0)
>>> op.pauli_rep
1.0 * X(0) + 1.0 * Y(0)
Extensive improvements have been made to the string representations of PennyLane operators, making them shorter and possible to copy-paste as valid PennyLane code. (#5116) (#5138)
>>> 0.5 * X(0)
0.5 * X(0)
>>> 0.5 * (X(0) + Y(1))
0.5 * (X(0) + Y(1))
Sums with many terms are broken up into multiple lines, but can still be copied back as valid code:
>>> 0.5 * (X(0) @ X(1)) + 0.7 * (X(1) @ X(2)) + 0.8 * (X(2) @ X(3))
(
0.5 * (X(0) @ X(1))
+ 0.7 * (X(1) @ X(2))
+ 0.8 * (X(2) @ X(3))
)
Linear combinations of operators and operator multiplication via Sum
and Prod
, respectively, have been updated to reach feature parity with Hamiltonian
and Tensor
, respectively. This should minimize the effort to port over any existing code. (#5070) (#5132) (#5133)
Updates include support for grouping via the pauli
module:
>>> obs = [X(0) @ Y(1), Z(0), Y(0) @ Z(1), Y(1)]
>>> qml.pauli.group_observables(obs)
[[Y(0) @ Z(1)], [X(0) @ Y(1), Y(1)], [Z(0)]]
A new default.clifford
device enables efficient simulation of large-scale Clifford circuits defined in PennyLane through the use of stim as a backend. (#4936) (#4954) (#5144)
Given a circuit with only Clifford gates, one can use this device to obtain the usual range of PennyLane measurements as well as the state represented in the Tableau form of Aaronson & Gottesman (2004):
import pennylane as qml
dev = qml.device("default.clifford", tableau=True)
@qml.qnode(dev)
def circuit():
qml.CNOT(wires=[0, 1])
qml.PauliX(wires=[1])
qml.ISWAP(wires=[0, 1])
qml.Hadamard(wires=[0])
return qml.state()
>>> circuit()
array([[0, 1, 1, 0, 0],
[1, 0, 1, 1, 1],
[0, 0, 0, 1, 0],
[1, 0, 0, 1, 1]])
The default.clifford
device also supports the PauliError
, DepolarizingChannel
, BitFlip
and PhaseFlip
noise channels when operating in finite-shot mode.
Vector-Jacobian products (VJPs) can result in faster computations when the output of your quantum Node has a low dimension. They can be enabled by setting device_vjp=True
when loading a QNode. In the next release of PennyLane, VJPs are planned to be used by default, when available.
In this release, we have unlocked:
Adjoint device VJPs can be used with jax.jacobian
, meaning that device_vjp=True
is always faster when using JAX with default.qubit
. (#4963)
PennyLane can now use lightning-provided VJPs. (#4914)
VJPs can be used with TensorFlow, though support has not yet been added for tf.Function
and Tensorflow Autograph. (#4676)
Measuring qml.probs
is now faster due to an optimization in converting samples to counts. (#5145)
The performance of circuit-cutting workloads with large numbers of generated tapes has been improved. (#5005)
Queueing (AnnotatedQueue
) has been removed from qml.cut_circuit
and qml.cut_circuit_mc
to improve performance for large workflows. (#5108)
A new function called qml.fermi.parity_transform
has been added for parity mapping of a fermionic Hamiltonian. (#4928)
It is now possible to transform a fermionic Hamiltonian to a qubit Hamiltonian with parity mapping.
import pennylane as qml
fermi_ham = qml.fermi.FermiWord({(0, 0) : '+', (1, 1) : '-'})
qubit_ham = qml.fermi.parity_transform(fermi_ham, n=6)
>>> print(qubit_ham)
-0.25j * Y(0) + (-0.25+0j) * (X(0) @ Z(1)) + (0.25+0j) * X(0) + 0.25j * (Y(0) @ Z(1))
The transform split_non_commuting
now accepts measurements of type probs
, sample
, and counts
, which accept both wires and observables. (#4972)
The efficiency of matrix calculations when an operator is symmetric over a given set of wires has been improved. (#3601)
The pennylane/math/quantum.py
module now has support for computing the minimum entropy of a density matrix. (#3959)
>>> x = [1, 0, 0, 1] / np.sqrt(2)
>>> x = qml.math.dm_from_state_vector(x)
>>> qml.math.min_entropy(x, indices=[0])
0.6931471805599455
A function called apply_operation
that applies operations to device-compatible states has been added to the new qutrit_mixed
module found in qml.devices
. (#5032)
A function called measure
has been added to the new qutrit_mixed
module found in qml.devices
that measures device-compatible states for a collection of measurement processes. (#5049)
A partial_trace
function has been added to qml.math
for taking the partial trace of matrices. (#5152)
The following capabilities have been added for Pauli arithmetic: (#4989) (#5001) (#5003) (#5017) (#5027) (#5018)
You can now multiply PauliWord
and PauliSentence
instances by scalars (e.g., 0.5 * PauliWord({0: "X"})
or 0.5 * PauliSentence({PauliWord({0: "X"}): 1.})
).
You can now intuitively add and subtract PauliWord
and PauliSentence
instances and scalars together (scalars are treated implicitly as multiples of the identity, I
). For example, ps1 + pw1 + 1.
for some Pauli word pw1 = PauliWord({0: "X", 1: "Y"})
and Pauli sentence ps1 = PauliSentence({pw1: 3.})
.
You can now element-wise multiply PauliWord
, PauliSentence
, and operators together with qml.dot
(e.g., qml.dot([0.5, -1.5, 2], [pw1, ps1, id_word])
with id_word = PauliWord({})
).
qml.matrix
now accepts PauliWord
and PauliSentence
instances (e.g., qml.matrix(PauliWord({0: "X"}))
).
It is now possible to compute commutators with Pauli operators natively with the new commutator
method.
>>> op1 = PauliWord({0: "X", 1: "X"})
>>> op2 = PauliWord({0: "Y"}) + PauliWord({1: "Y"})
>>> op1.commutator(op2) 2j * Z(0) @ X(1) + 2j * X(0) @ Z(1)
Composite operations (e.g., those made with qml.prod
and qml.sum
) and scalar-product operations convert Hamiltonian
and Tensor
operands to Sum
and Prod
types, respectively. This helps avoid the mixing of incompatible operator types. (#5031) (#5063)
qml.Identity()
can be initialized without wires. Measuring it is currently not possible, though. (#5106)
qml.dot
now returns a Sum
class even when all the coefficients match. (#5143)
qml.pauli.group_observables
now supports grouping Prod
and SProd
operators. (#5070)
The performance of converting a PauliSentence
to a Sum
has been improved. (#5141) (#5150)
Akin to qml.Hamiltonian
features, the coefficients and operators that make up composite operators formed via Sum
or Prod
can now be accessed with the terms()
method. (#5132) (#5133) (#5164)
>>> qml.operation.enable_new_opmath()
>>> op = X(0) @ (0.5 * X(1) + X(2))
>>> op.terms()
([0.5, 1.0],
[X(1) @ X(0),
X(2) @ X(0)])
String representations of ParametrizedHamiltonian
have been updated to match the style of other PL operators. (#5215)
The pl-device-test
suite is now compatible with the qml.devices.Device
interface. (#5229)
The QSVT
operation now determines its data
from the block encoding and projector operator data. (#5226) (#5248)
The BlockEncode
operator is now JIT-compatible with JAX. (#5110)
The qml.qsvt
function uses qml.GlobalPhase
instead of qml.exp
to define a global phase. (#5105)
The tests/ops/functions/conftest.py
test has been updated to ensure that all operator types are tested for validity. (#4978)
A new pennylane.workflow
module has been added. This module now contains qnode.py
, execution.py
, set_shots.py
, jacobian_products.py
, and the submodule interfaces
. (#5023)
A more informative error is now raised when calling adjoint_jacobian
with trainable state-prep operations. (#5026)
qml.workflow.get_transform_program
and qml.workflow.construct_batch
have been added to inspect the transform program and batch of tapes at different stages. (#5084)
All custom controlled operations such as CRX
, CZ
, CNOT
, ControlledPhaseShift
now inherit from ControlledOp
, giving them additional properties such as control_wire
and control_values
. Calling qml.ctrl
on RX
, RY
, RZ
, Rot
, and PhaseShift
with a single control wire will return gates of types CRX
, CRY
, etc. as opposed to a general Controlled
operator. (#5069) (#5199)
The CI will now fail if coverage data fails to upload to codecov. Previously, it would silently pass and the codecov check itself would never execute. (#5101)
qml.ctrl
called on operators with custom controlled versions will now return instances of the custom class, and it will flatten nested controlled operators to a single multi-controlled operation. For PauliX
, CNOT
, Toffoli
, and MultiControlledX
, calling qml.ctrl
will always resolve to the best option in CNOT
, Toffoli
, or MultiControlledX
depending on the number of control wires and control values. (#5125)
Unwanted warning filters have been removed from tests and no PennyLaneDeprecationWarning
s are being raised unexpectedly. (#5122)
New error tracking and propagation functionality has been added (#5115) (#5121)
The method map_batch_transform
has been replaced with the method _batch_transform
implemented in TransformDispatcher
. (#5212)
TransformDispatcher
can now dispatch onto a batch of tapes, making it easier to compose transforms when working in the tape paradigm. (#5163)
qml.ctrl
is now a simple wrapper that either calls PennyLane's built in create_controlled_op
or uses the Catalyst implementation. (#5247)
Controlled composite operations can now be decomposed using ZYZ rotations. (#5242)
New functions called qml.devices.modifiers.simulator_tracking
and qml.devices.modifiers.single_tape_support
have been added to add basic default behavior onto a device class. (#5200)
Passing additional arguments to a transform that decorates a QNode must now be done through the use of functools.partial
. (#5046)
qml.ExpvalCost
has been removed. Users should use qml.expval()
moving forward. (#5097)
Caching of executions is now turned off by default when max_diff == 1
, as the classical overhead cost outweighs the probability that duplicate circuits exists. (#5243)
The entry point convention registering compilers with PennyLane has changed. (#5140)
To allow for packages to register multiple compilers with PennyLane, the entry_points
convention under the designated group name pennylane.compilers
has been modified.
Previously, compilers would register qjit
(JIT decorator), ops
(compiler-specific operations), and context
(for tracing and program capture).
Now, compilers must register compiler_name.qjit
, compiler_name.ops
, and compiler_name.context
, where compiler_name
is replaced by the name of the provided compiler.
For more information, please see the documentation on adding compilers.
PennyLane source code is now compatible with the latest version of black
. (#5112) (#5119)
gradient_analysis_and_validation
has been renamed to find_and_validate_gradient_methods
. Instead of returning a list, it now returns a dictionary of gradient methods for each parameter index, and no longer mutates the tape. (#5035)
Multiplying two PauliWord
instances no longer returns a tuple (new_word, coeff)
but instead PauliSentence({new_word: coeff})
. The old behavior is still available with the private method PauliWord._matmul(other)
for faster processing. (#5045)
Observable.return_type
has been removed. Instead, you should inspect the type of the surrounding measurement process. (#5044)
ClassicalShadow.entropy()
no longer needs an atol
keyword as a better method to estimate entropies from approximate density matrix reconstructions (with potentially negative eigenvalues). (#5048)
Controlled operators with a custom controlled version decompose like how their controlled counterpart decomposes as opposed to decomposing into their controlled version. (#5069) (#5125)
For example:
>>> qml.ctrl(qml.RX(0.123, wires=1), control=0).decomposition()
[
RZ(1.5707963267948966, wires=[1]),
RY(0.0615, wires=[1]),
CNOT(wires=[0, 1]),
RY(-0.0615, wires=[1]),
CNOT(wires=[0, 1]),
RZ(-1.5707963267948966, wires=[1])
]
QuantumScript.is_sampled
and QuantumScript.all_sampled
have been removed. Users should now validate these properties manually. (#5072)
qml.transforms.one_qubit_decomposition
and qml.transforms.two_qubit_decomposition
have been removed. Instead, you should use qml.ops.one_qubit_decomposition
and qml.ops.two_qubit_decomposition
. (#5091)
Calling qml.matrix
without providing a wire_order
on objects where the wire order could be ambiguous now raises a warning. In the future, the wire_order
argument will be required in these cases. (#5039)
Operator.validate_subspace(subspace)
has been relocated to the qml.ops.qutrit.parametric_ops
module and will be removed from the Operator class in an upcoming release. (#5067)
Matrix and tensor products between PauliWord
and PauliSentence
instances are done using the @
operator, *
will be used only for scalar multiplication. Note also the breaking change that the product of two PauliWord
instances now returns a PauliSentence
instead of a tuple (new_word, coeff)
. (#4989) (#5054)
MeasurementProcess.name
and MeasurementProcess.data
are now deprecated, as they contain dummy values that are no longer needed. (#5047) (#5071) (#5076) (#5122)
qml.pauli.pauli_mult
and qml.pauli.pauli_mult_with_phase
are now deprecated. Instead, you should use qml.simplify(qml.prod(pauli_1, pauli_2))
to get the reduced operator. (#5057)
The private functions _pauli_mult
, _binary_matrix
and _get_pauli_map
from the pauli
module have been deprecated, as they are no longer used anywhere and the same functionality can be achieved using newer features in the pauli
module. (#5057)
Sum.ops
, Sum.coeffs
, Prod.ops
and Prod.coeffs
will be deprecated in the future. (#5164)
The module documentation for pennylane.tape
now explains the difference between QuantumTape
and QuantumScript
. (#5065)
A typo in a code example in the qml.transforms
API has been fixed. (#5014)
Documentation for qml.data
has been updated and now mentions a way to access the same dataset simultaneously from multiple environments. (#5029)
A clarification for the definition of argnum
added to gradient methods has been made. (#5035)
A typo in the code example for qml.qchem.dipole_of
has been fixed. (#5036)
A development guide on deprecations and removals has been added. (#5083)
A note about the eigenspectrum of second-quantized Hamiltonians has been added to qml.eigvals
. (#5095)
A warning about two mathematically equivalent Hamiltonians undergoing different time evolutions has been added to qml.TrotterProduct
and qml.ApproxTimeEvolution
. (#5137)
A reference to the paper that provides the image of the qml.QAOAEmbedding
template has been added. (#5130)
The docstring of qml.sample
has been updated to advise the use of single-shot expectations instead when differentiating a circuit. (#5237)
A quick start page has been added called "Importing Circuits". This explains how to import quantum circuits and operations defined outside of PennyLane. (#5281)
QubitChannel
can now be used with jitting. (#5288)
Fixed a bug in the matplotlib drawer where the colour of Barrier
did not match the requested style. (#5276)
qml.draw
and qml.draw_mpl
now apply all applied transforms before drawing. (#5277)
ctrl_decomp_zyz
is now differentiable. (#5198)
qml.ops.Pow.matrix()
is now differentiable with TensorFlow with integer exponents. (#5178)
The qml.MottonenStatePreparation
template has been updated to include a global phase operation. (#5166)
Fixed a queuing bug when using qml.prod
with a quantum function that queues a single operator. (#5170)
The qml.TrotterProduct
template has been updated to accept scalar products of operators as an input Hamiltonian. (#5073)
Fixed a bug where caching together with JIT compilation and broadcasted tapes yielded wrong results Operator.hash
now depends on the memory location, id
, of a JAX tracer instead of its string representation. (#3917)
qml.transforms.undo_swaps
can now work with operators with hyperparameters or nesting. (#5081)
qml.transforms.split_non_commuting
will now pass the original shots along. (#5081)
If argnum
is provided to a gradient transform, only the parameters specified in argnum
will have their gradient methods validated. (#5035)
StatePrep
operations expanded onto more wires are now compatible with backprop. (#5028)
qml.equal
works well with qml.Sum
operators when wire labels are a mix of integers and strings. (#5037)
The return value of Controlled.generator
now contains a projector that projects onto the correct subspace based on the control value specified. (#5068)
CosineWindow
no longer raises an unexpected error when used on a subset of wires at the beginning of a circuit. (#5080)
tf.function
now works with TensorSpec(shape=None)
by skipping batch size computation. (#5089)
PauliSentence.wires
no longer imposes a false order. (#5041)
qml.qchem.import_state
now applies the chemist-to-physicist sign convention when initializing a PennyLane state vector from classically pre-computed wavefunctions. That is, it interleaves spin-up/spin-down operators for the same spatial orbital index, as standard in PennyLane (instead of commuting all spin-up operators to the left, as is standard in quantum chemistry). (#5114)
Multi-wire controlled CNOT
and PhaseShift
are now be decomposed correctly. (#5125) (#5148)
draw_mpl
no longer raises an error when drawing a circuit containing an adjoint of a controlled operation. (#5149)
default.mixed
no longer throws ValueError
when applying a state vector that is not of type complex128
when used with tensorflow. (#5155)
ctrl_decomp_zyz
no longer raises a TypeError
if the rotation parameters are of type torch.Tensor
(#5183)
Comparing Prod
and Sum
objects now works regardless of nested structure with qml.equal
if the operators have a valid pauli_rep
property. (#5177)
Controlled GlobalPhase
with non-zero control wires no longer throws an error. (#5194)
A QNode
transformed with mitigate_with_zne
now accepts batch parameters. (#5195)
The matrix of an empty PauliSentence
instance is now correct (all-zeros). Further, matrices of empty PauliWord
and PauliSentence
instances can now be turned into matrices. (#5188)
PauliSentence
instances can handle matrix multiplication with PauliWord
instances. (#5208)
CompositeOp.eigendecomposition
is now JIT-compatible. (#5207)
QubitDensityMatrix
now works with JAX-JIT on the default.mixed
device. (#5203) (#5236)
When a QNode specifies diff_method="adjoint"
, default.qubit
no longer tries to decompose non-trainable operations with non-scalar parameters such as QubitUnitary
. (#5233)
The overwriting of the class names of I
, X
, Y
, and Z
no longer happens in the initialization after causing problems with datasets. This now happens globally. (#5252)
The adjoint_metric_tensor
transform now works with jax
. (#5271)
This release contains contributions from (in alphabetical order):
Abhishek Abhishek, Mikhail Andrenkov, Utkarsh Azad, Trenten Babcock, Gabriel Bottrill, Thomas Bromley, Astral Cai, Skylar Chan, Isaac De Vlugt, Diksha Dhawan, Lillian Frederiksen, Pietropaolo Frisoni, Eugenio Gigante, Diego Guala, David Ittah, Soran Jahangiri, Jacky Jiang, Korbinian Kottmann, Christina Lee, Xiaoran Li, Vincent Michaud-Rioux, Romain Moyard, Pablo Antonio Moreno Casares, Erick Ochoa Lopez, Lee J. O'Riordan, Mudit Pandey, Alex Preciado, Matthew Silverman, Jay Soni.
Published by albi3ro 9 months ago
This postfix release pins additional requirements in doc/requirements.txt
for building the website documentation. This allows the website to be rebuilt to show the "Getting involved" section.
Published by lillian542 9 months ago
It is now possible to return statistics of composite mid-circuit measurements. (#4888)
Mid-circuit measurement results can be composed using basic arithmetic operations and then statistics can be calculated by putting the result within a PennyLane measurement like qml.expval()
. For example:
import pennylane as qml
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circuit(phi, theta):
qml.RX(phi, wires=0)
m0 = qml.measure(wires=0)
qml.RY(theta, wires=1)
m1 = qml.measure(wires=1)
return qml.expval(~m0 + m1)
print(circuit(1.23, 4.56))
1.2430187928114291
Another option, for ease-of-use when using qml.sample()
, qml.probs()
, or qml.counts()
, is to provide a simple list of mid-circuit measurement results:
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circuit(phi, theta):
qml.RX(phi, wires=0)
m0 = qml.measure(wires=0)
qml.RY(theta, wires=1)
m1 = qml.measure(wires=1)
return qml.sample(op=[m0, m1])
print(circuit(1.23, 4.56, shots=5))
[[0 1]
[0 1]
[0 0]
[1 0]
[0 1]]
Composite mid-circuit measurement statistics are supported on default.qubit
and default.mixed
. To learn more about which measurements and arithmetic operators are supported, refer to the measurements page and the documentation for qml.measure.
Mid-circuit measurements can now be visualized with the text-based qml.draw()
and the graphical qml.draw_mpl()
methods. (#4775) (#4803) (#4832) (#4901) (#4850) (#4917) (#4930) (#4957)
Drawing of mid-circuit measurement capabilities including qubit reuse and reset, postselection, conditioning, and collecting statistics is now supported. Here is an all-encompassing example:
def circuit():
m0 = qml.measure(0, reset=True)
m1 = qml.measure(1, postselect=1)
qml.cond(m0 - m1 == 0, qml.S)(0)
m2 = qml.measure(1)
qml.cond(m0 + m1 == 2, qml.T)(0)
qml.cond(m2, qml.PauliX)(1)
The text-based drawer outputs:
>>> print(qml.draw(circuit)())
0: ──┤↗│ │0⟩────────S───────T────┤
1: ───║────────┤↗₁├──║──┤↗├──║──X─┤
╚═════════║════╬═══║═══╣ ║
╚════╩═══║═══╝ ║
╚══════╝
The graphical drawer outputs:
>>> print(qml.draw_mpl(circuit)())
Catalyst, our next-generation compilation framework, is now accessible within PennyLane, allowing you to more easily benefit from hybrid just-in-time (JIT) compilation.
To access these features, simply install pennylane-catalyst
:
pip install pennylane-catalyst
The qml.compiler module provides support for hybrid quantum-classical compilation. (#4692) (#4979)
Through the use of the qml.qjit
decorator, entire workflows can be JIT compiled — including both quantum and classical processing — down to a machine binary on first-function execution. Subsequent calls to the compiled function will execute the previously-compiled binary, resulting in significant performance improvements.
import pennylane as qml
dev = qml.device("lightning.qubit", wires=2)
@qml.qjit
@qml.qnode(dev)
def circuit(theta):
qml.Hadamard(wires=0)
qml.RX(theta, wires=1)
qml.CNOT(wires=[0,1])
return qml.expval(qml.PauliZ(wires=1))
>>> circuit(0.5) # the first call, compilation occurs here array(0.)
>>> circuit(0.5) # the precompiled quantum function is called
array(0.)
Currently, PennyLane supports the Catalyst hybrid compiler with the qml.qjit
decorator. A significant benefit of Catalyst is the ability to preserve complex control flow around quantum operations — such as if
statements and for
loops, and including measurement feedback — during compilation, while continuing to support end-to-end autodifferentiation.
The following functions can now be used with the qml.qjit
decorator: qml.grad
, qml.jacobian
, qml.vjp
, qml.jvp
, and qml.adjoint
. (#4709) (#4724) (#4725) (#4726)
When qml.grad
or qml.jacobian
are used with @qml.qjit
, they are patched to catalyst.grad and catalyst.jacobian, respectively.
dev = qml.device("lightning.qubit", wires=1)
@qml.qjit
def workflow(x):
@qml.qnode(dev)
def circuit(x):
qml.RX(np.pi * x[0], wires=0)
qml.RY(x[1], wires=0)
return qml.probs()
g = qml.jacobian(circuit)
return g(x)
>>> workflow(np.array([2.0, 1.0]))
array([[ 3.48786850e-16, -4.20735492e-01],
[-8.71967125e-17, 4.20735492e-01]])
JIT-compatible functionality for control flow has been added via qml.for_loop
, qml.while_loop
, and qml.cond
. (#4698)
qml.for_loop
and qml.while_loop
can be deployed as decorators on functions that are the body of the loop. The arguments to both follow typical conventions:
@qml.for_loop(lower_bound, upper_bound, step)
@qml.while_loop(cond_function)
Here is a concrete example with qml.for_loop
:
dev = qml.device("lightning.qubit", wires=1)
@qml.qjit
@qml.qnode(dev)
def circuit(n: int, x: float):
@qml.for_loop(0, n, 1)
def loop_rx(i, x):
# perform some work and update (some of) the arguments
qml.RX(x, wires=0)
# update the value of x for the next iteration
return jnp.sin(x)
# apply the for loop
final_x = loop_rx(x)
return qml.expval(qml.PauliZ(0)), final_x
>>> circuit(7, 1.6)
(array(0.97926626), array(0.55395718))
The new qml.clifford_t_decomposition()
transform provides an approximate breakdown of an input circuit into the Clifford+T gateset. Behind the scenes, this decomposition is enacted via the sk_decomposition()
function using the Solovay-Kitaev algorithm. (#4801) (#4802)
The Solovay-Kitaev algorithm approximately decomposes a quantum circuit into the Clifford+T gateset. To account for this, a desired total circuit decomposition error, epsilon
, must be specified when using qml.clifford_t_decomposition
:
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circuit():
qml.RX(1.1, 0)
return qml.state()
circuit = qml.clifford_t_decomposition(circuit, epsilon=0.1)
>>> print(qml.draw(circuit)())
0: ──T†──H──T†──H──T──H──T──H──T──H──T──H──T†──H──T†──T†──H──T†──H──T──H──T──H──T──H──T──H──T†──H
───T†──H──T──H──GlobalPhase(0.39)─┤
The resource requirements of this circuit can also be evaluated:
>>> with qml.Tracker(dev) as tracker:
... circuit()
>>> resources_lst = tracker.history["resources"]
>>> resources_lst[0]
wires: 1
gates: 34
depth: 34
shots: Shots(total=None)
gate_types: {'Adjoint(T)': 8, 'Hadamard': 16, 'T': 9, 'GlobalPhase': 1}
gate_sizes: {1: 33, 0: 1}
Iterative Quantum Phase Estimation is now available with qml.iterative_qpe
. (#4804)
The subroutine can be used similarly to mid-circuit measurements:
import pennylane as qml
dev = qml.device("default.qubit", shots=5)
@qml.qnode(dev)
def circuit():
# Initial state
qml.PauliX(wires=[0])
# Iterative QPE
measurements = qml.iterative_qpe(qml.RZ(2., wires=[0]), ancilla=[1], iters=3)
return [qml.sample(op=meas) for meas in measurements]
>>> print(circuit())
[array([0, 0, 0, 0, 0]), array([1, 0, 0, 0, 0]), array([0, 1, 1, 1, 1])]
The $i$-th element in the list refers to the 5 samples generated by the $i$-th measurement of the algorithm.
The +=
operand can now be used with a PauliSentence
, which has also provides a performance boost. (#4662)
The Approximate Quantum Fourier Transform (AQFT) is now available with qml.AQFT
. (#4715)
qml.draw
and qml.draw_mpl
now render operator IDs. (#4749)
The ID can be specified as a keyword argument when instantiating an operator:
>>> def circuit():
... qml.RX(0.123, id="data", wires=0)
>>> print(qml.draw(circuit)())
0: ──RX(0.12,"data")─┤
Non-parametric operators such as Barrier
, Snapshot
, and Wirecut
have been grouped together and moved to pennylane/ops/meta.py
. Additionally, the relevant tests have been organized and placed in a new file, tests/ops/test_meta.py
. (#4789)
The TRX
, TRY
, and TRZ
operators are now differentiable via backpropagation on default.qutrit
. (#4790)
The function qml.equal
now supports ControlledSequence
operators. (#4829)
XZX decomposition has been added to the list of supported single-qubit unitary decompositions. (#4862)
==
and !=
operands can now be used with TransformProgram
and TransformContainers
instances. (#4858)
A qutrit_mixed
module has been added to qml.devices
to store helper functions for a future qutrit mixed-state device. A function called create_initial_state
has been added to this module that creates device-compatible initial states. (#4861)
The function qml.Snapshot
now supports arbitrary state-based measurements (i.e., measurements of type StateMeasurement
). (#4876)
qml.equal
now supports the comparison of QuantumScript
and BasisRotation
objects. (#4902) (#4919)
The function qml.draw_mpl
now accept a keyword argument fig
to specify the output figure window. (#4956)
qml.AmplitudeEmbedding
now supports batching when used with Tensorflow. (#4818)
default.qubit
can now evolve already batched states with qml.pulse.ParametrizedEvolution
. (#4863)
qml.ArbitraryUnitary
now supports batching. (#4745)
Operator and tape batch sizes are evaluated lazily, helping run expensive computations less frequently and an issue with Tensorflow pre-computing batch sizes. (#4911)
Autograd, PyTorch, and JAX can now use vector-Jacobian products (VJPs) provided by the device from the new device API. If a device provides a VJP, this can be selected by providing device_vjp=True
to a QNode or qml.execute
. (#4935) (#4557) (#4654) (#4878) (#4841)
>>> dev = qml.device('default.qubit')
>>> @qml.qnode(dev, diff_method="adjoint", device_vjp=True)
>>> def circuit(x):
... qml.RX(x, wires=0)
... return qml.expval(qml.PauliZ(0))
>>> with dev.tracker:
... g = qml.grad(circuit)(qml.numpy.array(0.1))
>>> dev.tracker.totals
{'batches': 1, 'simulations': 1, 'executions': 1, 'vjp_batches': 1, 'vjps': 1}
>>> g
-0.09983341664682815
qml.expval
with large Hamiltonian
objects is now faster and has a significantly lower memory footprint (and constant with respect to the number of Hamiltonian
terms) when the Hamiltonian
is a PauliSentence
. This is due to the introduction of a specialized dot
method in the PauliSentence
class which performs PauliSentence
-state
products. (#4839)
default.qubit
no longer uses a dense matrix for MultiControlledX
for more than 8 operation wires. (#4673)
Some relevant Pytests have been updated to enable its use as a suite of benchmarks. (#4703)
default.qubit
now applies GroverOperator
faster by not using its matrix representation but a custom rule for apply_operation
. Also, the matrix representation of GroverOperator
now runs faster. (#4666)
A new pipeline to run benchmarks and plot graphs comparing with a fixed reference has been added. This pipeline will run on a schedule and can be activated on a PR with the label ci:run_benchmarks
. (#4741)
default.qubit
now supports adjoint differentiation for arbitrary diagonal state-based measurements. (#4865)
The benchmarks pipeline has been expanded to export all benchmark data to a single JSON file and a CSV file with runtimes. This includes all references and local benchmarks. (#4873)
qml.quantum_monte_carlo
and qml.simplify
now use the new transform system. (#4708) (#4949)
The formal requirement that type hinting be provided when using the qml.transform
decorator has been removed. Type hinting can still be used, but is now optional. Please use a type checker such as mypy if you wish to ensure types are being passed correctly. (#4942)
PennyLane now supports Python 3.12. (#4985)
SampleMeasurement
now has an optional method process_counts
for computing the measurement results from a counts dictionary. (#4941)
A new function called ops.functions.assert_valid
has been added for checking if an Operator
class is defined correctly. (#4764)
Shots
objects can now be multiplied by scalar values. (#4913)
GlobalPhase
now decomposes to nothing in case devices do not support global phases. (#4855)
Custom operations can now provide their matrix directly through the Operator.matrix()
method without needing to update the has_matrix
property. has_matrix
will now automatically be True
if Operator.matrix
is overridden, even if Operator.compute_matrix
is not. (#4844)
The logic for re-arranging states before returning them has been improved. (#4817)
When multiplying SparseHamiltonian
s by a scalar value, the result now stays as a SparseHamiltonian
. (#4828)
trainable_params
can now be set upon initialization of a QuantumScript
instead of having to set the parameter after initialization. (#4877)
default.qubit
now calculates the expectation value of Hermitian
operators in a differentiable manner. (#4866)
The rot
decomposition now has support for returning a global phase. (#4869)
The "pennylane_sketch"
MPL-drawer style has been added. This is the same as the "pennylane"
style, but with sketch-style lines. (#4880)
Operators now define a pauli_rep
property, an instance of PauliSentence
, defaulting to None
if the operator has not defined it (or has no definition in the Pauli basis). (#4915)
qml.ShotAdaptiveOptimizer
can now use a multinomial distribution for spreading shots across the terms of a Hamiltonian measured in a QNode. Note that this is equivalent to what can be done with qml.ExpvalCost
, but this is the preferred method because ExpvalCost
is deprecated. (#4896)
Decomposition of qml.PhaseShift
now uses qml.GlobalPhase
for retaining the global phase information. (#4657) (#4947)
qml.equal
for Controlled
operators no longer returns False
when equivalent but differently-ordered sets of control wires and control values are compared. (#4944)
All PennyLane Operator
subclasses are automatically tested by ops.functions.assert_valid
to ensure that they follow PennyLane Operator
standards. (#4922)
Probability measurements can now be calculated from a counts
dictionary with the addition of a process_counts
method in the ProbabilityMP
class. (#4952)
ClassicalShadow.entropy
now uses the algorithm outlined in 1106.5458 to project the approximate density matrix (with potentially negative eigenvalues) onto the closest valid density matrix. (#4959)
The ControlledSequence.compute_decomposition
default now decomposes the Pow
operators, improving compatibility with machine learning interfaces. (#4995)
The functions qml.transforms.one_qubit_decomposition
, qml.transforms.two_qubit_decomposition
, and qml.transforms.sk_decomposition
were moved to qml.ops.one_qubit_decomposition
, qml.ops.two_qubit_decomposition
, and qml.ops.sk_decomposition
, respectively. (#4906)
The function qml.transforms.classical_jacobian
has been moved to the gradients module and is now accessible as qml.gradients.classical_jacobian
. (#4900)
The transforms submodule qml.transforms.qcut
is now its own module: qml.qcut
. (#4819)
The decomposition of GroverOperator
now has an additional global phase operation. (#4666)
qml.cond
and the Conditional
operation have been moved from the transforms
folder to the ops/op_math
folder. qml.transforms.Conditional
will now be available as qml.ops.Conditional
. (#4860)
The prep
keyword argument has been removed from QuantumScript
and QuantumTape
. StatePrepBase
operations should be placed at the beginning of the ops
list instead. (#4756)
qml.gradients.pulse_generator
is now named qml.gradients.pulse_odegen
to adhere to paper naming conventions. (#4769)
Specifying control_values
passed to qml.ctrl
as a string is no longer supported. (#4816)
The rot
decomposition will now normalize its rotation angles to the range [0, 4pi]
for consistency (#4869)
QuantumScript.graph
is now built using tape.measurements
instead of tape.observables
because it depended on the now-deprecated Observable.return_type
property. (#4762)
The "pennylane"
MPL-drawer style now draws straight lines instead of sketch-style lines. (#4880)
The default value for the term_sampling
argument of ShotAdaptiveOptimizer
is now None
instead of "weighted_random_sampling"
. (#4896)
single_tape_transform
, batch_transform
, qfunc_transform
, and op_transform
are deprecated. Use the new qml.transform
function instead. (#4774)
Observable.return_type
is deprecated. Instead, you should inspect the type of the surrounding measurement process. (#4762) (#4798)
All deprecations now raise a qml.PennyLaneDeprecationWarning
instead of a UserWarning
. (#4814)
QuantumScript.is_sampled
and QuantumScript.all_sampled
are deprecated. Users should now validate these properties manually. (#4773)
With an algorithmic improvement to ClassicalShadow.entropy
, the keyword atol
becomes obsolete and will be removed in v0.35. (#4959)
Documentation for unitaries and operations' decompositions has been moved from qml.transforms
to qml.ops.ops_math
. (#4906)
Documentation for qml.metric_tensor
and qml.adjoint_metric_tensor
and qml.transforms.classical_jacobian
is now accessible via the gradients API page qml.gradients
in the documentation. (#4900)
Documentation for qml.specs
has been moved to the resource
module. (#4904)
Documentation for QCut has been moved to its own API page: qml.qcut
. (#4819)
The documentation page for qml.measurements
now links top-level accessible functions (e.g., qml.expval
) to their top-level pages rather than their module-level pages (e.g., qml.measurements.expval
). (#4750)
Information for the documentation of qml.matrix
about wire ordering has been added for using qml.matrix
on a QNode which uses a device with device.wires=None
. (#4874)
TransformDispatcher
now stops queuing when performing the transform when applying it to a qfunc. Only the output of the transform will be queued. (#4983)
qml.map_wires
now works properly with qml.cond
and qml.measure
. (#4884)
Pow
operators are now picklable. (#4966)
Finite differences and SPSA can now be used with tensorflow-autograph on setups that were seeing a bus error. (#4961)
qml.cond
no longer incorrectly queues operators used arguments. (#4948)
Attribute
objects now return False
instead of raising a TypeError
when checking if an object is inside the set. (#4933)
Fixed a bug where the parameter-shift rule of qml.ctrl(op)
was wrong if op
had a generator that has two or more eigenvalues and is stored as a SparseHamiltonian
. (#4899)
Fixed a bug where trainable parameters in the post-processing of finite-differences were incorrect for JAX when applying the transform directly on a QNode. (#4879)
qml.grad
and qml.jacobian
now explicitly raise errors if trainable parameters are integers. (#4836)
JAX-JIT now works with shot vectors. (#4772)
JAX can now differentiate a batch of circuits where one tape does not have trainable parameters. (#4837)
The decomposition of GroverOperator
now has the same global phase as its matrix. (#4666)
The tape.to_openqasm
method no longer mistakenly includes interface information in the parameter string when converting tapes using non-NumPy interfaces. (#4849)
qml.defer_measurements
now correctly transforms circuits when terminal measurements include wires used in mid-circuit measurements. (#4787)
Fixed a bug where the adjoint differentiation method would fail if an operation that has a parameter with grad_method=None
is present. (#4820)
MottonenStatePreparation
and BasisStatePreparation
now raise an error when decomposing a broadcasted state vector. (#4767)
Gradient transforms now work with overridden shot vectors and default.qubit
. (#4795)
Any ScalarSymbolicOp
, like Evolution
, now states that it has a matrix if the target is a Hamiltonian
. (#4768)
In default.qubit
, initial states are now initialized with the simulator's wire order, not the circuit's wire order. (#4781)
qml.compile
will now always decompose to expand_depth
, even if a target basis set is not specified. (#4800)
qml.transforms.transpile
can now handle measurements that are broadcasted onto all wires. (#4793)
Parametrized circuits whose operators do not act on all wires return PennyLane tensors instead of NumPy arrays, as expected. (#4811) (#4817)
qml.transforms.merge_amplitude_embedding
no longer depends on queuing, allowing it to work as expected with QNodes. (#4831)
qml.pow(op)
and qml.QubitUnitary.pow()
now also work with Tensorflow data raised to an integer power. (#4827)
The text drawer has been fixed to correctly label qml.qinfo
measurements, as well as qml.classical_shadow
qml.shadow_expval
. (#4803)
Removed an implicit assumption that an empty PauliSentence
gets treated as identity under multiplication. (#4887)
Using a CNOT
or PauliZ
operation with large batched states and the Tensorflow interface no longer raises an unexpected error. (#4889)
qml.map_wires
no longer fails when mapping nested quantum tapes. (#4901)
Conversion of circuits to openqasm now decomposes to a depth of 10, allowing support for operators requiring more than 2 iterations of decomposition, such as the ApproxTimeEvolution
gate. (#4951)
MPLDrawer
does not add the bonus space for classical wires when no classical wires are present. (#4987)
Projector
now works with parameter-broadcasting. (#4993) * The jax-jit interface can now be used with float32 mode. (#4990)
Keras models with a qnn.KerasLayer
no longer fail to save and load weights properly when they are named "weights". (#5008)
This release contains contributions from (in alphabetical order):
Guillermo Alonso, Ali Asadi, Utkarsh Azad, Gabriel Bottrill, Thomas Bromley, Astral Cai, Minh Chau, Isaac De Vlugt, Amintor Dusko, Pieter Eendebak, Lillian Frederiksen, Pietropaolo Frisoni, Josh Izaac, Juan Giraldo, Emiliano Godinez Ramirez, Ankit Khandelwal, Korbinian Kottmann, Christina Lee, Vincent Michaud-Rioux, Anurav Modak, Romain Moyard, Mudit Pandey, Matthew Silverman, Jay Soni, David Wierichs, Justin Woodring.
Published by timmysilv 12 months ago
Fix gradient performance regression due to expansion of VJP products. (#4806)
qml.defer_measurements
now correctly transforms circuits when terminal measurements include wires used in mid-circuit measurements. (#4787)
Any ScalarSymbolicOp
, like Evolution
, now states that it has a matrix if the target is a Hamiltonian
. (#4768)
In default.qubit
, initial states are now initialized with the simulator's wire order, not the circuit's wire order. (#4781)
This release contains contributions from (in alphabetical order):
Christina Lee, Lee James O'Riordan, Mudit Pandey
Published by mudit2812 12 months ago
It is now possible to request postselection on a mid-circuit measurement. (#4604)
This can be achieved by specifying the postselect
keyword argument in qml.measure
as either 0
or 1
, corresponding to the basis states.
import pennylane as qml
dev = qml.device("default.qubit")
@qml.qnode(dev, interface=None)
def circuit():
qml.Hadamard(wires=0)
qml.CNOT(wires=[0, 1])
qml.measure(0, postselect=1)
return qml.expval(qml.PauliZ(1)), qml.sample(wires=1)
This circuit prepares the $| \Phi^{+} \rangle$ Bell state and postselects on measuring $|1\rangle$ in wire 0
. The output of wire 1
is then also $|1\rangle$ at all times:
>>> circuit(shots=10)
(-1.0, array([1, 1, 1, 1, 1, 1]))
Note that the number of shots is less than the requested amount because we have thrown away the samples where $|0\rangle$ was measured in wire 0
.
Measurement statistics can now be collected for mid-circuit measurements. (#4544)
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circ(x, y):
qml.RX(x, wires=0)
qml.RY(y, wires=1)
m0 = qml.measure(1)
return qml.expval(qml.PauliZ(0)), qml.expval(m0), qml.sample(m0)
>>> circ(1.0, 2.0, shots=10000)
(0.5606, 0.7089, array([0, 1, 1, ..., 1, 1, 1]))
Support is provided for both finite-shot and analytic modes and devices default to using the deferred measurement principle to enact the mid-circuit measurements.
Higher-order Trotter-Suzuki methods are now easily accessible through a new operation called TrotterProduct
. (#4661)
Trotterization techniques are an affective route towards accurate and efficient Hamiltonian simulation. The Suzuki-Trotter product formula allows for the ability to express higher-order approximations to the matrix exponential of a Hamiltonian, and it is now available to use in PennyLane via the TrotterProduct
operation. Simply specify the order
of the approximation and the evolution time
.
coeffs = [0.25, 0.75]
ops = [qml.PauliX(0), qml.PauliZ(0)]
H = qml.dot(coeffs, ops)
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def circuit():
qml.Hadamard(0)
qml.TrotterProduct(H, time=2.4, order=2)
return qml.state()
>>> circuit()
[-0.13259524+0.59790098j 0. +0.j -0.13259524-0.77932754j 0. +0.j ]
Approximating matrix exponentiation with random product formulas, qDrift, is now available with the new QDrift
operation. (#4671)
As shown in 1811.08017, qDrift is a Markovian process that can provide a speedup in Hamiltonian simulation. At a high level, qDrift works by randomly sampling from the Hamiltonian terms with a probability that depends on the Hamiltonian coefficients. This method for Hamiltonian simulation is now ready to use in PennyLane with the QDrift
operator. Simply specify the evolution time
and the number of samples drawn from the Hamiltonian, n
:
coeffs = [0.25, 0.75]
ops = [qml.PauliX(0), qml.PauliZ(0)]
H = qml.dot(coeffs, ops)
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def circuit():
qml.Hadamard(0)
qml.QDrift(H, time=1.2, n = 10)
return qml.probs()
>>> circuit()
array([0.61814334, 0. , 0.38185666, 0. ])
A new operator called CosineWindow
has been added to prepare an initial state based on a cosine wave function. (#4683)
As outlined in 2110.09590, the cosine tapering window is part of a modification to quantum phase estimation that can provide a cubic improvement to the algorithm's error rate. Using CosineWindow
will prepare a state whose amplitudes follow a cosinusoidal distribution over the computational basis.
import matplotlib.pyplot as plt
dev = qml.device('default.qubit', wires=4)
@qml.qnode(dev)
def example_circuit():
qml.CosineWindow(wires=range(4))
return qml.state()
output = example_circuit()
plt.style.use("pennylane.drawer.plot")
plt.bar(range(len(output)), output)
plt.show()
Controlled gate sequences raised to decreasing powers, a sub-block in quantum phase estimation, can now be created with the new ControlledSequence
operator. (#4707)
To use ControlledSequence
, specify the controlled unitary operator and the control wires, control
:
dev = qml.device("default.qubit", wires = 4)
@qml.qnode(dev)
def circuit():
for i in range(3):
qml.Hadamard(wires = i)
qml.ControlledSequence(qml.RX(0.25, wires = 3), control = [0, 1, 2])
qml.adjoint(qml.QFT)(wires = range(3))
return qml.probs(wires = range(3))
>>> print(circuit())
[0.92059345 0.02637178 0.00729619 0.00423258 0.00360545 0.00423258 0.00729619 0.02637178]
default.qubit
now uses the new qml.devices.Device
API and functionality in qml.devices.qubit
. If you experience any issues with the updated default.qubit
, please let us know by posting an issue. The old version of the device is still accessible by the short name default.qubit.legacy
, or directly via qml.devices.DefaultQubitLegacy
. (#4594) (#4436) (#4620) (#4632)
This changeover has a number of benefits for default.qubit
, including:
The number of wires is now optional — simply having qml.device("default.qubit")
is valid! If wires are not provided at instantiation, the device automatically infers the required number of wires for each circuit provided for execution.
dev = qml.device("default.qubit")
@qml.qnode(dev)
def circuit():
qml.PauliZ(0)
qml.RZ(0.1, wires=1)
qml.Hadamard(2)
return qml.state()
>>> print(qml.draw(circuit)())
0: ──Z────────┤ State
1: ──RZ(0.10)─┤ State
2: ──H────────┤ State
default.qubit
is no longer silently swapped out with an interface-appropriate device when the backpropagation differentiation method is used. For example, consider:
import jax
dev = qml.device("default.qubit", wires=1)
@qml.qnode(dev, diff_method="backprop")
def f(x):
qml.RX(x, wires=0)
return qml.expval(qml.PauliZ(0))
f(jax.numpy.array(0.2))
In previous versions of PennyLane, the device will be swapped for the JAX equivalent:
>>> f.device
<DefaultQubitJax device (wires=1, shots=None) at 0x7f8c8bff50a0>
>>> f.device == dev
False
Now, default.qubit
can itself dispatch to all the interfaces in a backprop-compatible way and hence does not need to be swapped:
>>> f.device
<default.qubit device (wires=1) at 0x7f20d043b040>
>>> f.device == dev
True
A QNode that has been decorated with qjit
from PennyLane's Catalyst library for just-in-time hybrid compilation is now compatible with qml.draw
. (#4609)
import catalyst
@catalyst.qjit
@qml.qnode(qml.device("lightning.qubit", wires=3))
def circuit(x, y, z, c):
"""A quantum circuit on three wires."""
@catalyst.for_loop(0, c, 1)
def loop(i):
qml.Hadamard(wires=i)
qml.RX(x, wires=0)
loop()
qml.RY(y, wires=1)
qml.RZ(z, wires=2)
return qml.expval(qml.PauliZ(0))
draw = qml.draw(circuit, decimals=None)(1.234, 2.345, 3.456, 1)
>>> print(draw)
0: ──RX──H──┤ <Z>
1: ──H───RY─┤
2: ──RZ─────┤
MeasurementProcess
and QuantumScript
objects are now registered as JAX PyTrees. (#4607) (#4608)
It is now possible to JIT-compile functions with arguments that are a MeasurementProcess
or a QuantumScript
:
tape0 = qml.tape.QuantumTape([qml.RX(1.0, 0), qml.RY(0.5, 0)], [qml.expval(qml.PauliZ(0))])
dev = qml.device('lightning.qubit', wires=5)
execute_kwargs = {"device": dev, "gradient_fn": qml.gradients.param_shift, "interface":"jax"}
jitted_execute = jax.jit(qml.execute, static_argnames=execute_kwargs.keys())
jitted_execute((tape0, ), **execute_kwargs)
Computationally expensive functions in integrals.py
, electron_repulsion
and _hermite_coulomb
, have been modified to replace indexing with slicing for better compatibility with JAX. (#4685)
qml.qchem.import_state
has been extended to import more quantum chemistry wavefunctions, from MPS, DMRG and SHCI classical calculations performed with the Block2 and Dice libraries. #4523 #4524 #4626 #4634
Check out our how-to guide to learn more about how PennyLane integrates with your favourite quantum chemistry libraries.
The qchem fermionic_dipole
and particle_number
functions have been updated to use a FermiSentence
. The deprecated features for using tuples to represent fermionic operations are removed. (#4546) (#4556)
The tensor-network template qml.MPS
now supports changing the offset
between subsequent blocks for more flexibility. (#4531)
Builtin types support with qml.pauli_decompose
have been improved. (#4577)
AmplitudeEmbedding
now inherits from StatePrep
, allowing for it to not be decomposed when at the beginning of a circuit, thus behaving like StatePrep
. (#4583)
qml.cut_circuit
is now compatible with circuits that compute the expectation values of Hamiltonians with two or more terms. (#4642)
default.qubit
now tracks the number of equivalent qpu executions and total shots when the device is sampling. Note that "simulations"
denotes the number of simulation passes, whereas "executions"
denotes how many different computational bases need to be sampled in. Additionally, the new default.qubit
tracks the results of device.execute
. (#4628) (#4649)
DefaultQubit
can now accept a jax.random.PRNGKey
as a seed
to set the key for the JAX pseudo random number generator when using the JAX interface. This corresponds to the prng_key
on default.qubit.jax
in the old API. (#4596)
The JacobianProductCalculator
abstract base class and implementations TransformJacobianProducts
DeviceDerivatives
, and DeviceJacobianProducts
have been added to pennylane.interfaces.jacobian_products
. (#4435) (#4527) (#4637)
DefaultQubit
dispatches to a faster implementation for applying ParametrizedEvolution
to a state when it is more efficient to evolve the state than the operation matrix. (#4598) (#4620)
Wires can be provided to the new device API. (#4538) (#4562)
qml.sample()
in the new device API now returns a np.int64
array instead of np.bool8
. (#4539)
The new device API now has a repr()
method. (#4562)
DefaultQubit
now works as expected with measurement processes that don't specify wires. (#4580)
Various improvements to measurements have been made for feature parity between default.qubit.legacy
and the new DefaultQubit
. This includes not trying to squeeze batched CountsMP
results and implementing MutualInfoMP.map_wires
. (#4574)
devices.qubit.simulate
now accepts an interface keyword argument. If a QNode with DefaultQubit
specifies an interface, the result will be computed with that interface. (#4582)
ShotAdaptiveOptimizer
has been updated to pass shots to QNode executions instead of overriding device shots before execution. This makes it compatible with the new device API. (#4599)
pennylane.devices.preprocess
now offers the transforms decompose
, validate_observables
, validate_measurements
, validate_device_wires
, validate_multiprocessing_workers
, warn_about_trainable_observables
, and no_sampling
to assist in constructing devices under the new device API. (#4659)
Updated qml.device
, devices.preprocessing
and the tape_expand.set_decomposition
context manager to bring DefaultQubit
to feature parity with default.qubit.legacy
with regards to using custom decompositions. The DefaultQubit
device can now be included in a set_decomposition
context or initialized with a custom_decomps
dictionary, as well as a custom max_depth
for decomposition. (#4675)
The StateMP
measurement now accepts a wire order (e.g., a device wire order). The process_state
method will re-order the given state to go from the inputted wire-order to the process's wire-order. If the process's wire-order contains extra wires, it will assume those are in the zero-state. (#4570) (#4602)
Methods called add_transform
and insert_front_transform
have been added to TransformProgram
. (#4559)
Instances of the TransformProgram
class can now be added together. (#4549)
Transforms can now be applied to devices following the new device API. (#4667)
All gradient transforms have been updated to the new transform program system. (#4595)
Multi-controlled operations with a single-qubit special unitary target can now automatically decompose. (#4697)
pennylane.defer_measurements
will now exit early if the input does not contain mid circuit measurements. (#4659)
The density matrix aspects of StateMP
have been split into their own measurement process called DensityMatrixMP
. (#4558)
StateMeasurement.process_state
now assumes that the input is flat. ProbabilityMP.process_state
has been updated to reflect this assumption and avoid redundant reshaping. (#4602)
qml.exp
returns a more informative error message when decomposition is unavailable for non-unitary operators. (#4571)
Added qml.math.get_deep_interface
to get the interface of a scalar hidden deep in lists or tuples. (#4603)
Updated qml.math.ndim
and qml.math.shape
to work with built-in lists or tuples that contain interface-specific scalar dat (e.g., [(tf.Variable(1.1), tf.Variable(2.2))]
). (#4603)
When decomposing a unitary matrix with one_qubit_decomposition
and opting to include the GlobalPhase
in the decomposition, the phase is no longer cast to dtype=complex
. (#4653)
_qfunc_output
has been removed from QuantumScript
, as it is no longer necessary. There is still a _qfunc_output
property on QNode
instances. (#4651)
qml.data.load
properly handles parameters that come after 'full'
(#4663)
The qml.jordan_wigner
function has been modified to optionally remove the imaginary components of the computed qubit operator, if imaginary components are smaller than a threshold. (#4639)
qml.data.load
correctly performs a full download of the dataset after a partial download of the same dataset has already been performed. (#4681) * The performance of qml.data.load()
has been improved when partially loading a dataset (#4674)
Plots generated with the pennylane.drawer.plot
style of matplotlib.pyplot
now have black axis labels and are generated at a default DPI of 300. (#4690)
Shallow copies of the QNode
now also copy the execute_kwargs
and transform program. When applying a transform to a QNode
, the new qnode is only a shallow copy of the original and thus keeps the same device. (#4736)
QubitDevice
and CountsMP
are updated to disregard samples containing failed hardware measurements (record as np.NaN
) when tallying samples, rather than counting failed measurements as ground-state measurements, and to display qml.counts
coming from these hardware devices correctly. (#4739)
qml.defer_measurements
now raises an error if a transformed circuit measures qml.probs
, qml.sample
, or qml.counts
without any wires or observable, or if it measures qml.state
. (#4701)
The device test suite now converts device keyword arguments to integers or floats if possible. (#4640)
MeasurementProcess.eigvals()
now raises an EigvalsUndefinedError
if the measurement observable does not have eigenvalues. (#4544)
The __eq__
and __hash__
methods of Operator
and MeasurementProcess
no longer rely on the object's address in memory. Using ==
with operators and measurement processes will now behave the same as qml.equal
, and objects of the same type with the same data and hyperparameters will have the same hash. (#4536)
In the following scenario, the second and third code blocks show the previous and current behaviour of operator and measurement process equality, determined by ==
:
op1 = qml.PauliX(0)
op2 = qml.PauliX(0)
op3 = op1
Old behaviour:
>>> op1 == op2
False
>>> op1 == op3
True
New behaviour:
>>> op1 == op2
True
>>> op1 == op3
True
The __hash__
dunder method defines the hash of an object. The default hash of an object is determined by the objects memory address. However, the new hash is determined by the properties and attributes of operators and measurement processes. Consider the scenario below. The second and third code blocks show the previous and current behaviour.
op1 = qml.PauliX(0)
op2 = qml.PauliX(0)
Old behaviour:
>>> print({op1, op2})
{PauliX(wires=[0]), PauliX(wires=[0])}
New behaviour:
>>> print({op1, op2})
{PauliX(wires=[0])}
The old return type and associated functions qml.enable_return
and qml.disable_return
have been removed. (#4503)
The mode
keyword argument in QNode
has been removed. Please use grad_on_execution
instead. (#4503)
The CV observables qml.X
and qml.P
have been removed. Please use qml.QuadX
and qml.QuadP
instead. (#4533)
The sampler_seed
argument of qml.gradients.spsa_grad
has been removed. Instead, the sampler_rng
argument should be set, either to an integer value, which will be used to create a PRNG internally, or to a NumPy pseudo-random number generator (PRNG) created via np.random.default_rng(seed)
. (#4550)
The QuantumScript.set_parameters
method and the QuantumScript.data
setter have been removed. Please use QuantumScript.bind_new_parameters
instead. (#4548)
The method tape.unwrap()
and corresponding UnwrapTape
and Unwrap
classes have been removed. Instead of tape.unwrap()
, use qml.transforms.convert_to_numpy_parameters
. (#4535)
The RandomLayers.compute_decomposition
keyword argument ratio_imprivitive
has been changed to ratio_imprim
to match the call signature of the operation. (#4552)
The private TmpPauliRot
operator used for SpecialUnitary
no longer decomposes to nothing when the theta value is trainable. (#4585)
ProbabilityMP.marginal_prob
has been removed. Its contents have been moved into process_state
, which effectively just called marginal_prob
with np.abs(state) ** 2
. (#4602)
The following decorator syntax for transforms has been deprecated and will raise a warning: (#4457)
@transform_fn(**transform_kwargs)
@qml.qnode(dev)
def circuit():
...
If you are using a transform that has supporting transform_kwargs
, please call the transform directly using circuit = transform_fn(circuit, **transform_kwargs)
, or use functools.partial
:
@functools.partial(transform_fn, **transform_kwargs)
@qml.qnode(dev)
def circuit():
...
The prep
keyword argument in QuantumScript
has been deprecated and will be removed from QuantumScript
. StatePrepBase
operations should be placed at the beginning of the ops
list instead. (#4554)
qml.gradients.pulse_generator
has been renamed to qml.gradients.pulse_odegen
to adhere to paper naming conventions. During v0.33, pulse_generator
is still available but raises a warning. (#4633)
A warning section in the docstring for DefaultQubit
regarding the start method used in multiprocessing has been added. This may help users circumvent issues arising in Jupyter notebooks on macOS for example. (#4622)
Documentation improvements to the new device API have been made. The documentation now correctly states that interface-specific parameters are only passed to the device for backpropagation derivatives. (#4542)
Functions for qubit-simulation to the qml.devices
sub-page of the "Internal" section have been added. Note that these functions are unstable while device upgrades are underway. (#4555)
A documentation improvement to the usage example in the qml.QuantumMonteCarlo
page has been made. An integral was missing the differential $dx$. (#4593)
A documentation improvement for the use of the pennylane
style of qml.drawer
and the pennylane.drawer.plot
style of matplotlib.pyplot
has been made by clarifying the use of the default font. (#4690)
Jax jit now works when a probability measurement is broadcasted onto all wires. (#4742)
Fixed LocalHilbertSchmidt.compute_decomposition
so that the template can be used in a QNode. (#4719)
Fixes transforms.transpile
with arbitrary measurement processes. (#4732)
Providing work_wires=None
to qml.GroverOperator
no longer interprets None
as a wire. (#4668)
Fixed an issue where the __copy__
method of the qml.Select()
operator attempted to access un-initialized data. (#4551)
Fixed the skip_first
option in expand_tape_state_prep
. (#4564)
convert_to_numpy_parameters
now uses qml.ops.functions.bind_new_parameters
. This reinitializes the operation and makes sure everything references the new NumPy parameters. (#4540)
tf.function
no longer breaks ProbabilityMP.process_state
, which is needed by new devices. (#4470)
Fixed unit tests for qml.qchem.mol_data
. (#4591)
Fixed ProbabilityMP.process_state
so that it allows for proper Autograph compilation. Without this, decorating a QNode that returns an expval
with tf.function
would fail when computing the expectation. (#4590)
The torch.nn.Module
properties are now accessible on a pennylane.qnn.TorchLayer
. (#4611)
qml.math.take
with Pytorch now returns tensor[..., indices]
when the user requests the last axis (axis=-1
). Without the fix, it would wrongly return tensor[indices]
. (#4605)
Ensured the logging TRACE
level works with gradient-free execution. (#4669)
This release contains contributions from (in alphabetical order):
Guillermo Alonso, Utkarsh Azad, Thomas Bromley, Isaac De Vlugt, Jack Brown, Stepan Fomichev, Joana Fraxanet, Diego Guala, Soran Jahangiri, Edward Jiang, Korbinian Kottmann, Ivana Kurečić Christina Lee, Lillian M. A. Frederiksen, Vincent Michaud-Rioux, Romain Moyard, Daniel F. Nino, Lee James O'Riordan, Mudit Pandey, Matthew Silverman, Jay Soni.
Published by albi3ro about 1 year ago
This release changes doc/requirements.txt
to upgrade jax
, jaxlib
, and pin ml-dtypes
.
Published by mudit2812 about 1 year ago
It is now possible to encode an operator A
into a quantum circuit by decomposing it into a linear combination of unitaries using PREP (qml.StatePrep) and SELECT (qml.Select) routines. (#4431) (#4437) (#4444) (#4450) (#4506) (#4526)
Consider an operator A
composed of a linear combination of Pauli terms:
>>> A = qml.PauliX(2) + 2 * qml.PauliY(2) + 3 * qml.PauliZ(2)
A decomposable block-encoding circuit can be created:
def block_encode(A, control_wires):
probs = A.coeffs / np.sum(A.coeffs)
state = np.pad(np.sqrt(probs, dtype=complex), (0, 1))
unitaries = A.ops
qml.StatePrep(state, wires=control_wires)
qml.Select(unitaries, control=control_wires)
qml.adjoint(qml.StatePrep)(state, wires=control_wires)
>>> print(qml.draw(block_encode, show_matrices=False)(A, control_wires=[0, 1]))
0: ─╭|Ψ⟩─╭Select─╭|Ψ⟩†─┤
1: ─╰|Ψ⟩─├Select─╰|Ψ⟩†─┤
2: ──────╰Select───────┤
This circuit can be used as a building block within a larger QNode to perform algorithms such as QSVT and Hamiltonian simulation.
Decomposing a Hermitian matrix into a linear combination of Pauli words via qml.pauli_decompose
is now faster and differentiable. (#4395) (#4479) (#4493)
def find_coeffs(p):
mat = np.array([[3, p], [p, 3]])
A = qml.pauli_decompose(mat)
return A.coeffs
>>> import jax
>>> from jax import numpy as np
>>> jax.jacobian(find_coeffs)(np.array(2.))
Array([0., 1.], dtype=float32, weak_type=True)
Python-native logging can now be enabled with qml.logging.enable_logging()
. (#4377) (#4383)
Consider the following code that is contained in my_code.py
:
import pennylane as qml
qml.logging.enable_logging() # enables logging
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def f(x):
qml.RX(x, wires=0)
return qml.state()
f(0.5)
Executing my_code.py
with logging enabled will detail every step in PennyLane's pipeline that gets used to run your code.
$ python my_code.py
[1967-02-13 15:18:38,591][DEBUG][<PID 8881:MainProcess>] - pennylane.qnode.__init__()::"Creating QNode(func=<function f at 0x7faf2a6fbaf0>, device=<DefaultQubit device (wires=2, shots=None) at 0x7faf2a689b50>, interface=auto, diff_method=best, expansion_strategy=gradient, max_expansion=10, grad_on_execution=best, mode=None, cache=True, cachesize=10000, max_diff=1, gradient_kwargs={}"
...
Additional logging configuration settings can be specified by modifying the contents of the logging configuration file, which can be located by running qml.logging.config_path()
. Follow our logging docs page for more details!
Input states obtained from advanced quantum chemistry calculations can be used in a circuit. (#4427) (#4433) (#4461) (#4476) (#4505)
Quantum chemistry calculations rely on an initial state that is typically selected to be the trivial Hartree-Fock state. For molecules with a complicated electronic structure, using initial states obtained from affordable post-Hartree-Fock calculations helps to improve the efficiency of the quantum simulations. These calculations can be done with external quantum chemistry libraries such as PySCF.
It is now possible to import a PySCF solver object in PennyLane and extract the corresponding wave function in the form of a state vector that can be directly used in a circuit. First, perform your classical quantum chemistry calculations and then use the qml.qchem.import_state function to import the solver object and return a state vector.
>>> from pyscf import gto, scf, ci
>>> mol = gto.M(atom=[['H', (0, 0, 0)], ['H', (0,0,0.71)]], basis='sto6g')
>>> myhf = scf.UHF(mol).run()
>>> myci = ci.UCISD(myhf).run()
>>> wf_cisd = qml.qchem.import_state(myci, tol=1e-1)
>>> print(wf_cisd)
[ 0. +0.j 0. +0.j 0. +0.j 0.1066467 +0.j
1. +0.j 0. +0.j 0. +0.j 0. +0.j
2. +0.j 0. +0.j 0. +0.j 0. +0.j
-0.99429698+0.j 0. +0.j 0. +0.j 0. +0.j]
The state vector can be implemented in a circuit using qml.StatePrep
.
>>> dev = qml.device('default.qubit', wires=4)
>>> @qml.qnode(dev)
... def circuit():
... qml.StatePrep(wf_cisd, wires=range(4))
... return qml.state()
>>> print(circuit())
[ 0. +0.j 0. +0.j 0. +0.j 0.1066467 +0.j
1. +0.j 0. +0.j 0. +0.j 0. +0.j
2. +0.j 0. +0.j 0. +0.j 0. +0.j
-0.99429698+0.j 0. +0.j 0. +0.j 0. +0.j]
The currently supported post-Hartree-Fock methods are RCISD, UCISD, RCCSD, and UCCSD which denote restricted (R) and unrestricted (U) configuration interaction (CI) and coupled cluster (CC) calculations with single and double (SD) excitations.
PennyLane now allows you to define circuits that reuse a qubit after a mid-circuit measurement has taken place. Optionally, the wire can also be reset to the $|0\rangle$ state. (#4402) (#4432)
Post-measurement reset can be activated by setting reset=True
when calling qml.measure. In this version of PennyLane, executing circuits with qubit reuse will result in the defer_measurements transform being applied. This transform replaces each reused wire with an additional qubit. However, future releases of PennyLane will explore device-level support for qubit reuse without consuming additional qubits.
Qubit reuse and reset is also fully differentiable:
dev = qml.device("default.qubit", wires=4)
@qml.qnode(dev)
def circuit(p):
qml.RX(p, wires=0)
m = qml.measure(0, reset=True)
qml.cond(m, qml.Hadamard)(1)
qml.RX(p, wires=0)
m = qml.measure(0)
qml.cond(m, qml.Hadamard)(1)
return qml.expval(qml.PauliZ(1))
>>> jax.grad(circuit)(0.4)
Array(-0.35867804, dtype=float32, weak_type=True)
You can read more about mid-circuit measurements in the documentation, and stay tuned for more mid-circuit measurement features in the next few releases!
Circuit drawings and plots can now be created following a PennyLane style. (#3950)
The qml.draw_mpl
function accepts a style='pennylane'
argument to create PennyLane themed circuit diagrams:
def circuit(x, z):
qml.QFT(wires=(0,1,2,3))
qml.Toffoli(wires=(0,1,2))
qml.CSWAP(wires=(0,2,3))
qml.RX(x, wires=0)
qml.CRZ(z, wires=(3,0))
return qml.expval(qml.PauliZ(0))
qml.draw_mpl(circuit, style="pennylane")(1, 1)
PennyLane-styled plots can also be drawn by passing "pennylane.drawer.plot"
to Matplotlib's plt.style.use
function:
import matplotlib.pyplot as plt
plt.style.use("pennylane.drawer.plot")
for i in range(3):
plt.plot(np.random.rand(10))
If the font Quicksand Bold isn't available, an available default font is used instead.
Any class inheriting from Operator
is now automatically registered as a pytree with JAX. This unlocks the ability to jit functions of Operator
. (#4458)
>>> op = qml.adjoint(qml.RX(1.0, wires=0))
>>> jax.jit(qml.matrix)(op)
Array([[0.87758255-0.j , 0. +0.47942555j],
[0. +0.47942555j, 0.87758255-0.j ]], dtype=complex64, weak_type=True)
>>> jax.tree_util.tree_map(lambda x: x+1, op)
Adjoint(RX(2.0, wires=[0]))
All Operator
objects now define Operator._flatten
and Operator._unflatten
methods that separate trainable from untrainable components. These methods will be used in serialization and pytree registration. Custom operations may need an update to ensure compatibility with new PennyLane features. (#4483) (#4314)
The QuantumScript
class now has a bind_new_parameters
method that allows creation of new QuantumScript
objects with the provided parameters. (#4345)
The qml.gradients
module no longer mutates operators in-place for any gradient transforms. Instead, operators that need to be mutated are copied with new parameters. (#4220)
PennyLane no longer directly relies on Operator.__eq__
. (#4398)
qml.equal
no longer raises errors when operators or measurements of different types are compared. Instead, it returns False
. (#4315)
Transform programs are now integrated with the QNode. (#4404)
def null_postprocessing(results: qml.typing.ResultBatch) -> qml.typing.Result:
return results[0]
@qml.transforms.core.transform
def scale_shots(tape: qml.tape.QuantumTape, shot_scaling) -> (Tuple[qml.tape.QuantumTape], Callable):
new_shots = tape.shots.total_shots * shot_scaling
new_tape = qml.tape.QuantumScript(tape.operations, tape.measurements, shots=new_shots)
return (new_tape, ), null_postprocessing
dev = qml.devices.experimental.DefaultQubit2()
@partial(scale_shots, shot_scaling=2)
@qml.qnode(dev, interface=None)
def circuit():
return qml.sample(wires=0)
>>> circuit(shots=1)
array([False, False])
Transform Programs, qml.transforms.core.TransformProgram
, can now be called on a batch of circuits and return a new batch of circuits and a single post processing function. (#4364)
TransformDispatcher
now allows registration of custom QNode transforms. (#4466)
QNode transforms in qml.qinfo
now support custom wire labels. #4331
qml.transforms.adjoint_metric_tensor
now uses the simulation tools in qml.devices.qubit
instead of private methods of qml.devices.DefaultQubit
. (#4456)
Auxiliary wires and device wires are now treated the same way in qml.transforms.metric_tensor
as in qml.gradients.hadamard_grad
. All valid wire input formats for aux_wire
are supported. (#4328)
The experimental device interface has been integrated with the QNode for JAX, JAX-JIT, TensorFlow and PyTorch. (#4323) (#4352) (#4392) (#4393)
The experimental DefaultQubit2
device now supports computing VJPs and JVPs using the adjoint method. (#4374)
New functions called adjoint_jvp
and adjoint_vjp
that compute the JVP and VJP of a tape using the adjoint method have been added to qml.devices.qubit.adjoint_jacobian
(#4358)
DefaultQubit2
now accepts a max_workers
argument which controls multiprocessing. A ProcessPoolExecutor
executes tapes asynchronously using a pool of at most max_workers
processes. If max_workers
is None
or not given, only the current process executes tapes. If you experience any issue, say using JAX, TensorFlow, Torch, try setting max_workers
to None
. (#4319) (#4425)
qml.devices.experimental.Device
now accepts a shots keyword argument and has a shots
property. This property is only used to set defaults for a workflow, and does not directly influence the number of shots used in executions or derivatives. (#4388)
expand_fn()
for DefaultQubit2
has been updated to decompose StatePrep
operations present in the middle of a circuit. (#4444)
If no seed is specified on initialization with DefaultQubit2
, the local random number generator will be seeded from NumPy's global random number generator. (#4394)
pennylane/interfaces
has been refactored. The execute_fn
passed to the machine learning framework boundaries is now responsible for converting parameters to NumPy. The gradients module can now handle TensorFlow parameters, but gradient tapes now retain the original dtype
instead of converting to float64
. This may cause instability with finite-difference differentiation and float32
parameters. The machine learning boundary functions are now uncoupled from their legacy counterparts. (#4415)
qml.interfaces.set_shots
now accepts a Shots
object as well as int
's and tuples of int
's. (#4388)
Readability improvements and stylistic changes have been made to pennylane/interfaces/jax_jit_tuple.py
(#4379)
A HardwareHamiltonian
can now be summed with int
or float
objects. A sequence of HardwareHamiltonian
s can now be summed via the builtin sum
. (#4343)
qml.pulse.transmon_drive
has been updated in accordance with 1904.06560. In particular, the functional form has been changed from $\Omega(t)(\cos(\omega_d t + \phi) X - \sin(\omega_d t + \phi) Y)$ to $\Omega(t) \sin(\omega_d t + \phi) Y$. (#4418) (#4465) (#4478) (#4418)
The qchem
module has been upgraded to use the fermionic operators of the fermi
module. #4336 #4521
The calculation of Sum
, Prod
, SProd
, PauliWord
, and PauliSentence
sparse matrices are orders of magnitude faster. (#4475) (#4272) (#4411)
A function called qml.math.fidelity_statevector
that computes the fidelity between two state vectors has been added. (#4322)
qml.ctrl(qml.PauliX)
returns a CNOT
, Toffoli
, or MultiControlledX
operation instead of Controlled(PauliX)
. (#4339)
When given a callable, qml.ctrl
now does its custom pre-processing on all queued operators from the callable. (#4370)
The qchem
functions primitive_norm
and contracted_norm
have been modified to be compatible with higher versions of SciPy. The private function _fac2
for computing double factorials has also been added. #4321
tape_expand
now uses Operator.decomposition
instead of Operator.expand
in order to make more performant choices. (#4355)
CI now runs tests with TensorFlow 2.13.0 (#4472)
All tests in CI and pre-commit hooks now enable linting. (#4335)
The default label for a StatePrepBase
operator is now |Ψ⟩
. (#4340)
Device.default_expand_fn()
has been updated to decompose qml.StatePrep
operations present in the middle of a provided circuit. (#4437)
QNode.construct
has been updated to only apply the qml.defer_measurements
transform if the device does not natively support mid-circuit measurements. (#4516)
The application of the qml.defer_measurements
transform has been moved from QNode.construct
to qml.Device.batch_transform
to allow more fine-grain control over when defer_measurements
should be used. (#4432)
The label for ParametrizedEvolution
can display parameters with the requested format as set by the kwarg decimals
. Array-like parameters are displayed in the same format as matrices and stored in the cache. (#4151)
Applying gradient transforms to broadcasted/batched tapes has been deactivated until it is consistently supported for QNodes as well. (#4480)
Gradient transforms no longer implicitly cast float32
parameters to float64
. Finite difference differentiation with float32
parameters may no longer give accurate results. (#4415)
The do_queue
keyword argument in qml.operation.Operator
has been removed. Instead of setting do_queue=False
, use the qml.QueuingManager.stop_recording()
context. (#4317)
Operator.expand
now uses the output of Operator.decomposition
instead of what it queues. (#4355)
The gradients module no longer needs shot information passed to it explicitly, as the shots are on the tapes. (#4448)
qml.StatePrep
has been renamed to qml.StatePrepBase
and qml.QubitStateVector
has been renamed to qml.StatePrep
. qml.operation.StatePrep
and qml.QubitStateVector
are still accessible. (#4450)
Support for Python 3.8 has been dropped. (#4453)
MeasurementValue
's signature has been updated to accept a list of MidMeasureMP
's rather than a list of their IDs. (#4446)
The grouping_type
and grouping_method
keyword arguments have been removed from qchem.molecular_hamiltonian
. (#4301)
zyz_decomposition
and xyx_decomposition
have been removed. Use one_qubit_decomposition
instead. (#4301)
LieAlgebraOptimizer
has been removed. Use RiemannianGradientOptimizer
instead. (#4301)
Operation.base_name
has been removed. (#4301)
QuantumScript.name
has been removed. (#4301)
qml.math.reduced_dm
has been removed. Use qml.math.reduce_dm
or qml.math.reduce_statevector
instead. (#4301)
The qml.specs
dictionary no longer supports direct key access to certain keys. (#4301)
Instead, these quantities can be accessed as fields of the new Resources
object saved under specs_dict["resources"]
:
num_operations
is no longer supported, use specs_dict["resources"].num_gates
num_used_wires
is no longer supported, use specs_dict["resources"].num_wires
gate_types
is no longer supported, use specs_dict["resources"].gate_types
gate_sizes
is no longer supported, use specs_dict["resources"].gate_sizes
depth
is no longer supported, use specs_dict["resources"].depth
qml.math.purity
, qml.math.vn_entropy
, qml.math.mutual_info
, qml.math.fidelity
, qml.math.relative_entropy
, and qml.math.max_entropy
no longer support state vectors as input. (#4322)
The private QuantumScript._prep
list has been removed, and prep operations now go into the _ops
list. (#4485)
qml.enable_return
and qml.disable_return
have been deprecated. Please avoid calling disable_return
, as the old return system has been deprecated along with these switch functions. (#4316)
qml.qchem.jordan_wigner
has been deprecated. Use qml.jordan_wigner
instead. List input to define the fermionic operator has also been deprecated; the fermionic operators in the qml.fermi
module should be used instead. (#4332)
The qml.RandomLayers.compute_decomposition
keyword argument ratio_imprimitive
will be changed to ratio_imprim
to match the call signature of the operation. (#4314)
The CV observables qml.X
and qml.P
have been deprecated. Use qml.QuadX
and qml.QuadP
instead. (#4330)
The method tape.unwrap()
and corresponding UnwrapTape
and Unwrap
classes have been deprecated. Use convert_to_numpy_parameters
instead. (#4344)
The mode
keyword argument in QNode has been deprecated, as it was only used in the old return system (which has also been deprecated). Please use grad_on_execution
instead. (#4316)
The QuantumScript.set_parameters
method and the QuantumScript.data
setter have been deprecated. Please use QuantumScript.bind_new_parameters
instead. (#4346)
The __eq__
and __hash__
dunder methods of Operator
and MeasurementProcess
will now raise warnings to reflect upcoming changes to operator and measurement process equality and hashing. (#4144) (#4454) (#4489) (#4498)
The sampler_seed
argument of qml.gradients.spsa_grad
has been deprecated, along with a bug fix of the seed-setting behaviour. Instead, the sampler_rng
argument should be set, either to an integer value, which will be used to create a PRNG internally or to a NumPy pseudo-random number generator created via np.random.default_rng(seed)
. (#4165)
The qml.pulse.transmon_interaction
and qml.pulse.transmon_drive
documentation has been updated. (#4327)
qml.ApproxTimeEvolution.compute_decomposition()
now has a code example. (#4354)
The documentation for qml.devices.experimental.Device
has been improved to clarify some aspects of its use. (#4391)
Input types and sources for operators in qml.import_operator
are specified. (#4476)
qml.Projector
is pickle-able again. (#4452)
_copy_and_shift_params
does not cast or convert integral types, just relying on +
and *
's casting rules in this case. (#4477)
Sparse matrix calculations of SProd
s containing a Tensor
are now allowed. When using Tensor.sparse_matrix()
, it is recommended to use the wire_order
keyword argument over wires
. (#4424)
op.adjoint
has been replaced with qml.adjoint
in QNSPSAOptimizer
. (#4421)
jax.ad
(deprecated) has been replaced by jax.interpreters.ad
. (#4403)
metric_tensor
stops accidentally catching errors that stem from flawed wires assignments in the original circuit, leading to recursion errors. (#4328)
A warning is now raised if control indicators are hidden when calling qml.draw_mpl
(#4295)
qml.qinfo.purity
now produces correct results with custom wire labels. (#4331)
default.qutrit
now supports all qutrit operations used with qml.adjoint
. (#4348)
The observable data of qml.GellMann
now includes its index, allowing correct comparison between instances of qml.GellMann
, as well as Hamiltonians and Tensors containing qml.GellMann
. (#4366)
qml.transforms.merge_amplitude_embedding
now works correctly when the AmplitudeEmbedding
s have a batch dimension. (#4353)
The jordan_wigner
function has been modified to work with Hamiltonians built with an active space. (#4372)
When a style
option is not provided, qml.draw_mpl
uses the current style set from qml.drawer.use_style
instead of black_white
. (#4357)
qml.devices.qubit.preprocess.validate_and_expand_adjoint
no longer sets the trainable parameters of the expanded tape. (#4365)
qml.default_expand_fn
now selectively expands operations or measurements allowing more operations to be executed in circuits when measuring non-qwc Hamiltonians. (#4401)
qml.ControlledQubitUnitary
no longer reports has_decomposition
as True
when it does not really have a decomposition. (#4407)
qml.transforms.split_non_commuting
now correctly works on tapes containing both expval
and var
measurements. (#4426)
Subtracting a Prod
from another operator now works as expected. (#4441)
The sampler_seed
argument of qml.gradients.spsa_grad
has been changed to sampler_rng
. One can either provide an integer, which will be used to create a PRNG internally. Previously, this lead to the same direction being sampled, when num_directions
is greater than 1. Alternatively, one can provide a NumPy PRNG, which allows reproducibly calling spsa_grad
without getting the same results every time. (#4165) (#4482)
qml.math.get_dtype_name
now works with autograd array boxes. (#4494)
The backprop gradient of qml.math.fidelity
is now correct. (#4380)
This release contains contributions from (in alphabetical order):
Utkarsh Azad,
Thomas Bromley,
Isaac De Vlugt,
Amintor Dusko,
Stepan Fomichev,
Lillian M. A. Frederiksen,
Soran Jahangiri,
Edward Jiang,
Korbinian Kottmann,
Ivana Kurečić,
Christina Lee,
Vincent Michaud-Rioux,
Romain Moyard,
Lee James O'Riordan,
Mudit Pandey,
Borja Requena,
Matthew Silverman,
Jay Soni,
David Wierichs,
Frederik Wilde.
Published by eddddddy about 1 year ago
data.Dataset
now uses HDF5 instead of dill for serialization. (#4097)
The qchem
functions primitive_norm
and contracted_norm
are modified to be compatible with higher versions of scipy. (#4321)
This release contains contributions from (in alphabetical order):
Utkarsh Azad, Jack Brown, Diego Guala, Soran Jahangiri, Matthew Silverman
Published by timmysilv over 1 year ago
Fermionic operators and arithmetic are now available. (#4191) (#4195) (#4200) (#4201) (#4209) (#4229) (#4253) (#4255) (#4262) (#4278)
There are a couple of ways to create fermionic operators with this new feature:
qml.FermiC
and qml.FermiA
: the fermionic creation and annihilation operators, respectively. These operators are defined by passing the index of the orbital that the fermionic operator acts on. For instance, the operators a⁺(0)
and a(3)
are respectively constructed as
>>> qml.FermiC(0)
a⁺(0)
>>> qml.FermiA(3)
a(3)
These operators can be composed with (*
) and linearly combined with (+
and -
) other Fermi operators to create arbitrary fermionic Hamiltonians. Multiplying several Fermi operators together creates an operator that we call a Fermi word:
>>> word = qml.FermiC(0) * qml.FermiA(0) * qml.FermiC(3) * qml.FermiA(3)
>>> word
a⁺(0) a(0) a⁺(3) a(3)
Fermi words can be linearly combined to create a fermionic operator that we call a Fermi sentence:
>>> sentence = 1.2 * word - 0.345 * qml.FermiC(3) * qml.FermiA(3)
>>> sentence
1.2 * a⁺(0) a(0) a⁺(3) a(3)
- 0.345 * a⁺(3) a(3)
via qml.fermi.from_string: create a fermionic operator that represents multiple creation and annihilation operators being multiplied by each other (a Fermi word).
>>> qml.fermi.from_string('0+ 1- 0+ 1-')
a⁺(0) a(1) a⁺(0) a(1)
>>> qml.fermi.from_string('0^ 1 0^ 1')
a⁺(0) a(1) a⁺(0) a(1)
Fermi words created with from_string
can also be linearly combined to create a Fermi sentence:
>>> word1 = qml.fermi.from_string('0+ 0- 3+ 3-')
>>> word2 = qml.fermi.from_string('3+ 3-')
>>> sentence = 1.2 * word1 + 0.345 * word2
>>> sentence
1.2 * a⁺(0) a(0) a⁺(3) a(3)
+ 0.345 * a⁺(3) a(3)
Additionally, any fermionic operator, be it a single fermionic creation/annihilation operator, a Fermi word, or a Fermi sentence, can be mapped to the qubit basis by using qml.jordan_wigner:
>>> qml.jordan_wigner(sentence)
((0.4725+0j)*(Identity(wires=[0]))) + ((-0.4725+0j)*(PauliZ(wires=[3]))) + ((-0.3+0j)*(PauliZ(wires=[0]))) + ((0.3+0j)*(PauliZ(wires=[0]) @ PauliZ(wires=[3])))
Learn how to create fermionic Hamiltonians describing some simple chemical systems by checking out our fermionic operators demo!
PennyLane's Tracker now monitors the resource requirements of circuits being executed by the device. (#4045) (#4110)
Suppose we have a workflow that involves executing circuits with different qubit numbers. We can obtain the resource requirements as a function of the number of qubits by executing the workflow with the Tracker
context:
dev = qml.device("default.qubit", wires=4)
@qml.qnode(dev)
def circuit(n_wires):
for i in range(n_wires):
qml.Hadamard(i)
return qml.probs(range(n_wires))
with qml.Tracker(dev) as tracker:
for i in range(1, 5):
circuit(i)
The resource requirements of individual circuits can then be inspected as follows:
>>> resources = tracker.history["resources"]
>>> resources[0]
wires: 1
gates: 1
depth: 1
shots: Shots(total=None)
gate_types:
{'Hadamard': 1}
gate_sizes:
{1: 1}
>>> [r.num_wires for r in resources]
[1, 2, 3, 4]
Moreover, it is possible to predict the resource requirements without evaluating circuits using the null.qubit
device, which follows the standard execution pipeline but returns numeric zeros. Consider the following workflow that takes the gradient of a 50
-qubit circuit:
n_wires = 50
dev = qml.device("null.qubit", wires=n_wires)
weight_shape = qml.StronglyEntanglingLayers.shape(2, n_wires)
weights = np.random.random(weight_shape, requires_grad=True)
@qml.qnode(dev, diff_method="parameter-shift")
def circuit(weights):
qml.StronglyEntanglingLayers(weights, wires=range(n_wires))
return qml.expval(qml.PauliZ(0))
with qml.Tracker(dev) as tracker:
qml.grad(circuit)(weights)
The tracker can be inspected to extract resource requirements without requiring a 50-qubit circuit run:
>>> tracker.totals
{'executions': 451, 'batches': 2, 'batch_len': 451}
>>> tracker.history["resources"][0]
wires: 50
gates: 200
depth: 77
shots: Shots(total=None)
gate_types:
{'Rot': 100, 'CNOT': 100}
gate_sizes:
{1: 100, 2: 100}
Custom operations can now be constructed that solely define resource requirements — an explicit decomposition or matrix representation is not needed. (#4033)
PennyLane is now able to estimate the total resource requirements of circuits that include one or more of these operations, allowing you to estimate requirements for high-level algorithms composed of abstract subroutines.
These operations can be defined by inheriting from ResourcesOperation and overriding the resources()
method to return an appropriate Resources object:
class CustomOp(qml.resource.ResourcesOperation):
def resources(self):
n = len(self.wires)
r = qml.resource.Resources(
num_wires=n,
num_gates=n ** 2,
depth=5,
)
return r
>>> wires = [0, 1, 2]
>>> c = CustomOp(wires)
>>> c.resources()
wires: 3
gates: 9
depth: 5
shots: Shots(total=None)
gate_types:
{}
gate_sizes:
{}
A quantum circuit that contains CustomOp
can be created and inspected using qml.specs:
dev = qml.device("default.qubit", wires=wires)
@qml.qnode(dev)
def circ():
qml.PauliZ(wires=0)
CustomOp(wires)
return qml.state()
>>> specs = qml.specs(circ)()
>>> specs["resources"].depth
6
ParametrizedHamiltonian now has an improved string representation. (#4176)
>>> def f1(p, t): return p[0] * jnp.sin(p[1] * t)
>>> def f2(p, t): return p * t
>>> coeffs = [2., f1, f2]
>>> observables = [qml.PauliX(0), qml.PauliY(0), qml.PauliZ(0)]
>>> qml.dot(coeffs, observables)
(2.0*(PauliX(wires=[0])))
+ (f1(params_0, t)*(PauliY(wires=[0])))
+ (f2(params_1, t)*(PauliZ(wires=[0])))
The quantum information module now supports trace distance. (#4181)
Two cases are enabled for calculating the trace distance:
A QNode transform via qml.qinfo.trace_distance:
dev = qml.device('default.qubit', wires=2)
@qml.qnode(dev)
def circuit(param):
qml.RY(param, wires=0)
qml.CNOT(wires=[0, 1])
return qml.state()
>>> trace_distance_circuit = qml.qinfo.trace_distance(circuit, circuit, wires0=[0], wires1=[0])
>>> x, y = np.array(0.4), np.array(0.6)
>>> trace_distance_circuit((x,), (y,))
0.047862689546603415
Flexible post-processing via qml.math.trace_distance:
>>> rho = np.array([[0.3, 0], [0, 0.7]])
>>> sigma = np.array([[0.5, 0], [0, 0.5]])
>>> qml.math.trace_distance(rho, sigma)
0.19999999999999998
It is now possible to prepare qutrit basis states with qml.QutritBasisState. (#4185)
wires = range(2)
dev = qml.device("default.qutrit", wires=wires)
@qml.qnode(dev)
def qutrit_circuit():
qml.QutritBasisState([1, 1], wires=wires)
qml.TAdd(wires=wires)
return qml.probs(wires=1)
>>> qutrit_circuit()
array([0., 0., 1.])
A new transform called one_qubit_decomposition has been added to provide a unified interface for decompositions of a single-qubit unitary matrix into sequences of X, Y, and Z rotations. All decompositions simplify the rotations angles to be between 0
and 4
pi. (#4210) (#4246)
>>> from pennylane.transforms import one_qubit_decomposition
>>> U = np.array([[-0.28829348-0.78829734j, 0.30364367+0.45085995j],
... [ 0.53396245-0.10177564j, 0.76279558-0.35024096j]])
>>> one_qubit_decomposition(U, 0, "ZYZ")
[RZ(tensor(12.32427531, requires_grad=True), wires=[0]),
RY(tensor(1.14938178, requires_grad=True), wires=[0]),
RZ(tensor(1.73305815, requires_grad=True), wires=[0])]
>>> one_qubit_decomposition(U, 0, "XYX", return_global_phase=True)
[RX(tensor(10.84535137, requires_grad=True), wires=[0]),
RY(tensor(1.39749741, requires_grad=True), wires=[0]),
RX(tensor(0.45246584, requires_grad=True), wires=[0]),
(0.38469215914523336-0.9230449299422961j)*(Identity(wires=[0]))]
The has_unitary_generator
attribute in qml.ops.qubit.attributes
no longer contains operators with non-unitary generators. (#4183)
PennyLane Docker builds have been updated to include the latest plugins and interface versions. (#4178)
The stochastic parameter-shift gradient method can now be used with hardware-compatible Hamiltonians. (#4132) (#4215)
This new feature generalizes the stochastic parameter-shift gradient transform for pulses (stoch_pulse_grad
) to support Hermitian generating terms beyond just Pauli words in pulse Hamiltonians, which makes it hardware-compatible.
A new differentiation method called qml.gradients.pulse_generator is available, which combines classical processing with the parameter-shift rule for multivariate gates to differentiate pulse programs. Access it in your pulse programs by setting diff_method=qml.gradients.pulse_generator
. (#4160)
qml.pulse.ParametrizedEvolution
now uses batched compressed sparse row (BCSR
) format. This allows for computing Jacobians of the unitary directly even when dense=False
. (#4126)
def U(params):
H = jnp.polyval * qml.PauliZ(0) # time dependent Hamiltonian
Um = qml.evolve(H, dense=False)(params, t=10.)
return qml.matrix(Um)
params = jnp.array([[0.5]], dtype=complex)
jac = jax.jacobian(U, holomorphic=True)(params)
The TorchLayer
and KerasLayer
integrations with torch.nn
and Keras
have been upgraded. Consider the following TorchLayer
:
n_qubits = 2
dev = qml.device("default.qubit", wires=n_qubits)
@qml.qnode(dev)
def qnode(inputs, weights):
qml.AngleEmbedding(inputs, wires=range(n_qubits))
qml.BasicEntanglerLayers(weights, wires=range(n_qubits))
return [qml.expval(qml.PauliZ(wires=i)) for i in range(n_qubits)]
n_layers = 6
weight_shapes = {"weights": (n_layers, n_qubits)}
qlayer = qml.qnn.TorchLayer(qnode, weight_shapes)
The following features are now available:
Native support for parameter broadcasting. (#4131)
>>> batch_size = 10
>>> inputs = torch.rand((batch_size, n_qubits))
>>> qlayer(inputs)
>>> dev.num_executions == 1
True
The ability to draw a TorchLayer
and KerasLayer
using qml.draw()
and qml.draw_mpl()
. (#4197)
>>> print(qml.draw(qlayer, show_matrices=False)(inputs))
0: ─╭AngleEmbedding(M0)─╭BasicEntanglerLayers(M1)─┤ <Z>
1: ─╰AngleEmbedding(M0)─╰BasicEntanglerLayers(M1)─┤ <Z>
Support for KerasLayer
model saving and clearer instructions on TorchLayer
model saving. (#4149) (#4158)
>>> torch.save(qlayer.state_dict(), "weights.pt") # Saving
>>> qlayer.load_state_dict(torch.load("weights.pt")) # Loading
>>> qlayer.eval()
Hybrid models containing KerasLayer
or TorchLayer
objects can also be saved and loaded.
qml.Projector
now accepts a state vector representation, which enables the creation of projectors in any basis. (#4192)
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def circuit(state):
return qml.expval(qml.Projector(state, wires=[0, 1]))
zero_state = [0, 0]
plusplus_state = np.array([1, 1, 1, 1]) / 2
>>> circuit(zero_state)
tensor(1., requires_grad=True)
>>> circuit(plusplus_state)
tensor(0.25, requires_grad=True)
Three qutrit rotation operators have been added that are analogous to RX
, RY
, and RZ
:
qml.TRX
: an X rotationqml.TRY
: a Y rotationqml.TRZ
: a Z rotationQutrit devices now support parameter-shift differentiation. (#2845)
qchem.molecular_hamiltonian()
, qchem.qubit_observable()
, qchem.import_operator()
, and qchem.dipole_moment()
now return an arithmetic operator if enable_new_opmath()
is active.
(#4138) (#4159) (#4189) (#4204)
Non-cubic lattice support for all electron resource estimation has been added. (#3956)
The qchem.molecular_hamiltonian()
function has been upgraded to support custom wires for constructing differentiable Hamiltonians. The zero imaginary component of the Hamiltonian coefficients have been removed. (#4050) (#4094)
Jordan-Wigner transforms that cache Pauli gate objects have been accelerated. (#4046)
An error is now raised by qchem.molecular_hamiltonian
when the dhf
method is used for an open-shell system. This duplicates a similar error in qchem.Molecule
but makes it clear that the pyscf
backend can be used for open-shell calculations. (#4058)
Updated various qubit tapering methods to support operator arithmetic. (#4252)
The new device interface has been integrated with qml.execute
for autograd, backpropagation, and no differentiation. (#3903)
Support for adjoint differentiation has been added to the DefaultQubit2
device. (#4037)
A new function called measure_with_samples
that returns a sample-based measurement result given a state has been added. (#4083) (#4093) (#4162) (#4254)
DefaultQubit2.preprocess
now returns a new ExecutionConfig
object with decisions for gradient_method
, use_device_gradient
, and grad_on_execution
. (#4102)
Support for sample-based measurements has been added to the DefaultQubit2
device. (#4105) (#4114) (#4133) (#4172)
The DefaultQubit2
device now has a seed
keyword argument. (#4120)
Added a dense
keyword to ParametrizedEvolution
that allows forcing dense or sparse matrices. (#4079) (#4095) (#4285)
Adds the Type variables pennylane.typing.Result
and pennylane.typing.ResultBatch
for type hinting the result of an execution. (#4018)
qml.devices.ExecutionConfig
no longer has a shots
property, as it is now on the QuantumScript
. It now has a use_device_gradient
property. ExecutionConfig.grad_on_execution = None
indicates a request for "best"
, instead of a string. (#4102)
The new device interface for Jax has been integrated with qml.execute
. (#4137)
The new device interface is now integrated with qml.execute
for Tensorflow. (#4169)
The experimental device DefaultQubit2
now supports qml.Snapshot
. (#4193)
The experimental device interface is integrated with the QNode
. (#4196)
The new device interface in integrated with qml.execute
for Torch. (#4257)
QuantumScript
now has a shots
property, allowing shots to be tied to executions instead of devices. (#4067) (#4103) (#4106) (#4112)
Several Python built-in functions are now properly defined for instances of the Shots
class.
print
: printing Shots
instances is now human-readablestr
: converting Shots
instances to human-readable strings==
: equating two different Shots
instanceshash
: obtaining the hash values of Shots
instancesqml.devices.ExecutionConfig
no longer has a shots
property, as it is now on the QuantumScript
. It now has a use_device_gradient
property. ExecutionConfig.grad_on_execution = None
indicates a request for "best"
instead of a string. (#4102)
QuantumScript.shots
has been integrated with QNodes so that shots are placed on the QuantumScript
during QNode
construction. (#4110)
The gradients
module has been updated to use the new Shots
object internally (#4152)
qml.prod
now accepts a single quantum function input for creating new Prod
operators. (#4011)
DiagonalQubitUnitary
now decomposes into RZ
, IsingZZ
and MultiRZ
gates instead of a QubitUnitary
operation with a dense matrix. (#4035)
All objects being queued in an AnnotatedQueue
are now wrapped so that AnnotatedQueue
is not dependent on the has of any operators or measurement processes. (#4087)
A dense
keyword to ParametrizedEvolution
that allows forcing dense or sparse matrices has been added. (#4079) (#4095)
Added a new function qml.ops.functions.bind_new_parameters
that creates a copy of an operator with new parameters without mutating the original operator. (#4113) (#4256)
qml.CY
has been moved from qml.ops.qubit.non_parametric_ops
to qml.ops.op_math.controlled_ops
and now inherits from qml.ops.op_math.ControlledOp
. (#4116)
qml.CZ
now inherits from the ControlledOp
class and supports exponentiation to arbitrary powers with pow
, which is no longer limited to integers. It also supports sparse_matrix
and decomposition
representations. (#4117)
The construction of the Pauli representation for the Sum
class is now faster. (#4142)
qml.drawer.drawable_layers.drawable_layers
and qml.CircuitGraph
have been updated to not rely on Operator
equality or hash to work correctly. (#4143)
A transform dispatcher and program have been added. (#4109) (#4187)
Reduced density matrix functionality has been added via qml.math.reduce_dm
and qml.math.reduce_statevector
. Both functions have broadcasting support. (#4173)
The following functions in qml.qinfo
now support parameter broadcasting:
reduced_dm
purity
vn_entropy
mutual_info
fidelity
relative_entropy
trace_distance
The following functions in qml.math
now support parameter broadcasting:
purity
vn_entropy
mutual_info
fidelity
relative_entropy
max_entropy
sqrt_matrix
pulse.ParametrizedEvolution
now raises an error if the number of input parameters does not match the number of parametrized coefficients in the ParametrizedHamiltonian
that generates it. An exception is made for HardwareHamiltonian
s which are not checked. (#4216)
The default value for the show_matrices
keyword argument in all drawing methods is now True
. This allows for quick insights into broadcasted tapes, for example. (#3920)
Type variables for qml.typing.Result
and qml.typing.ResultBatch
have been added for type hinting the result of an execution. (#4108)
The Jax-JIT interface now uses symbolic zeros to determine trainable parameters. (#4075)
A new function called pauli.pauli_word_prefactor()
that extracts the prefactor for a given Pauli word has been added. (#4164)
Variable-length argument lists of functions and methods in some docstrings is now more clear. (#4242)
qml.drawer.drawable_layers.drawable_layers
and qml.CircuitGraph
have been updated to not rely on Operator
equality or hash to work correctly. (#4143)
Drawing mid-circuit measurements connected by classical control signals to conditional operations is now possible. (#4228)
The autograd interface now submits all required tapes in a single batch on the backward pass. (#4245)
The default value for the show_matrices
keyword argument in all drawing methods is now True
. This allows for quick insights into broadcasted tapes, for example. (#3920)
DiagonalQubitUnitary
now decomposes into RZ
, IsingZZ
, and MultiRZ
gates rather than a QubitUnitary
. (#4035)
Jax trainable parameters are now Tracer
instead of JVPTracer
. It is not always the right definition for the JIT interface, but we update them in the custom JVP using symbolic zeros. (#4075)
The experimental Device interface qml.devices.experimental.Device
now requires that the preprocess
method also returns an ExecutionConfig
object. This allows the device to choose what "best"
means for various hyperparameters like gradient_method
and grad_on_execution
. (#4007) (#4102)
Gradient transforms with Jax no longer support argnum
. Use argnums
instead. (#4076)
qml.collections
, qml.op_sum
, and qml.utils.sparse_hamiltonian
have been removed. (#4071)
The pennylane.transforms.qcut
module now uses (op, id(op))
as nodes in directed multigraphs that are used within the circuit cutting workflow instead of op
. This change removes the dependency of the module on the hash of operators. (#4227)
Operator.data
now returns a tuple
instead of a list
. (#4222)
The pulse differentiation methods, pulse_generator
and stoch_pulse_grad
, now raise an error when they are applied to a QNode directly. Instead, use differentiation via a JAX entry point (jax.grad
, jax.jacobian
, ...). (#4241)
LieAlgebraOptimizer
has been renamed to RiemannianGradientOptimizer
. (#4153)
Operation.base_name
has been deprecated. Please use Operation.name
or type(op).__name__
instead.
QuantumScript
's name
keyword argument and property have been deprecated. This also affects QuantumTape
and OperationRecorder
. (#4141)
The qml.grouping
module has been removed. Its functionality has been reorganized in the qml.pauli
module.
The public methods of DefaultQubit
are pending changes to follow the new device API, as used in DefaultQubit2
. Warnings have been added to the docstrings to reflect this. (#4145)
qml.math.reduced_dm
has been deprecated. Please use qml.math.reduce_dm
or qml.math.reduce_statevector
instead. (#4173)
qml.math.purity
, qml.math.vn_entropy
, qml.math.mutual_info
, qml.math.fidelity
, qml.math.relative_entropy
, and qml.math.max_entropy
no longer support state vectors as input. Please call qml.math.dm_from_state_vector
on the input before passing to any of these functions. (#4186)
The do_queue
keyword argument in qml.operation.Operator
has been deprecated. Instead of setting do_queue=False
, use the qml.QueuingManager.stop_recording()
context. (#4148)
zyz_decomposition
and xyx_decomposition
are now deprecated in favour of one_qubit_decomposition
. (#4230)
The documentation is updated to construct QuantumTape
upon initialization instead of with queuing. (#4243)
The docstring for qml.ops.op_math.Pow.__new__
is now complete and it has been updated along with qml.ops.op_math.Adjoint.__new__
. (#4231)
The docstring for qml.grad
now states that it should be used with the Autograd interface only. (#4202)
The description of mult
in the qchem.Molecule
docstring now correctly states the value of mult
that is supported. (#4058)
Fixed adjoint jacobian results with grad_on_execution=False
in the JAX-JIT interface. (#4217)
Fixed the matrix of SProd
when the coefficient is tensorflow and the target matrix is not complex128
. (#4249)
Fixed a bug where stoch_pulse_grad
would ignore prefactors of rescaled Pauli words in the generating terms of a pulse Hamiltonian. (#4156)
Fixed a bug where the wire ordering of the wires
argument to qml.density_matrix
was not taken into account. (#4072)
A patch in interfaces/autograd.py
that checks for the strawberryfields.gbs
device has been removed. That device is pinned to PennyLane <= v0.29.0, so that patch is no longer necessary. (#4089)
qml.pauli.are_identical_pauli_words
now treats all identities as equal. Identity terms on Hamiltonians with non-standard wire orders are no longer eliminated. (#4161)
qml.pauli_sentence()
is now compatible with empty Hamiltonians qml.Hamiltonian([], [])
. (#4171)
Fixed a bug with Jax where executing multiple tapes with gradient_fn="device"
would fail. (#4190)
A more meaningful error message is raised when broadcasting with adjoint differentiation on DefaultQubit
. (#4203)
The has_unitary_generator
attribute in qml.ops.qubit.attributes
no longer contains operators with non-unitary generators. (#4183)
Fixed a bug where op = qml.qsvt()
was incorrect up to a global phase when using convention="Wx""
and qml.matrix(op)
. (#4214)
Fixed a buggy calculation of the angle in xyx_decomposition
that causes it to give an incorrect decomposition. An if
conditional was intended to prevent divide by zero errors, but the division was by the sine of the argument. So, any multiple of $\pi$ should trigger the conditional, but it was only checking if the argument was 0. Example: qml.Rot(2.3, 2.3, 2.3)
(#4210)
Fixed bug that caused ShotAdaptiveOptimizer
to truncate dimensions of parameter-distributed shots during optimization. (#4240)
Sum
observables can now have trainable parameters. (#4251) (#4275)
This release contains contributions from (in alphabetical order):
Venkatakrishnan AnushKrishna,
Utkarsh Azad,
Thomas Bromley,
Isaac De Vlugt,
Lillian M. A. Frederiksen,
Emiliano Godinez Ramirez
Nikhil Harle
Soran Jahangiri,
Edward Jiang,
Korbinian Kottmann,
Christina Lee,
Vincent Michaud-Rioux,
Romain Moyard,
Tristan Nemoz,
Mudit Pandey,
Manul Patel,
Borja Requena,
Modjtaba Shokrian-Zini,
Mainak Roy,
Matthew Silverman,
Jay Soni,
Edward Thomas,
David Wierichs,
Frederik Wilde.
Published by albi3ro over 1 year ago
Support for loading time-dependent Hamiltonians that are compatible with quantum hardware has been added, making it possible to load a Hamiltonian that describes an ensemble of Rydberg atoms or a collection of transmon qubits. (#3749) (#3911) (#3930) (#3936) (#3966) (#3987) (#4021) (#4040)
Rydberg atoms are the foundational unit for neutral atom quantum computing. A Rydberg-system Hamiltonian can be constructed from a drive term
qml.pulse.rydberg_drive
— and anqml.pulse.rydberg_interaction
:from jax import numpy as jnp
atom_coordinates = [[0, 0], [0, 4], [4, 0], [4, 4]]
wires = [0, 1, 2, 3]
amplitude = lambda p, t: p * jnp.sin(jnp.pi * t)
phase = jnp.pi / 2
detuning = 3 * jnp.pi / 4
H_d = qml.pulse.rydberg_drive(amplitude, phase, detuning, wires)
H_i = qml.pulse.rydberg_interaction(atom_coordinates, wires)
H = H_d + H_i
The time-dependent Hamiltonian H
can be used in a PennyLane pulse-level differentiable circuit:
dev = qml.device("default.qubit.jax", wires=wires)
@qml.qnode(dev, interface="jax")
def circuit(params):
qml.evolve(H)(params, t=[0, 10])
return qml.expval(qml.PauliZ(0))
>>> params = jnp.array([2.4])
>>> circuit(params)
Array(0.6316659, dtype=float32)
>>> import jax
>>> jax.grad(circuit)(params)
Array([1.3116529], dtype=float32)
The qml.pulse page contains additional details. Check out our release blog post for demonstration of how to perform the execution on actual hardware!
A pulse-level circuit can now be differentiated using a stochastic parameter-shift method. (#3780) (#3900) (#4000) (#4004)
The new qml.gradient.stoch_pulse_grad differentiation method unlocks stochastic-parameter-shift differentiation for pulse-level circuits. The current version of this new method is restricted to Hamiltonians composed of parametrized Pauli words, but future updates to extend to parametrized Pauli sentences can allow this method to be compatible with hardware-based systems such as an ensemble of Rydberg atoms.
This method can be activated by setting diff_method
to qml.gradient.stoch_pulse_grad:
>>> dev = qml.device("default.qubit.jax", wires=2)
>>> sin = lambda p, t: jax.numpy.sin(p * t)
>>> ZZ = qml.PauliZ(0) @ qml.PauliZ(1)
>>> H = 0.5 * qml.PauliX(0) + qml.pulse.constant * ZZ + sin * qml.PauliX(1)
>>> @qml.qnode(dev, interface="jax", diff_method=qml.gradients.stoch_pulse_grad)
>>> def ansatz(params):
... qml.evolve(H)(params, (0.2, 1.))
... return qml.expval(qml.PauliY(1))
>>> params = [jax.numpy.array(0.4), jax.numpy.array(1.3)]
>>> jax.grad(ansatz)(params)
[Array(0.16921353, dtype=float32, weak_type=True),
Array(-0.2537478, dtype=float32, weak_type=True)]
PennyLane now supports the quantum singular value transformation (QSVT), which describes how a quantum circuit can be constructed to apply a polynomial transformation to the singular values of an input matrix. (#3756) (#3757) (#3758) (#3905) (#3909) (#3926) (#4023)
Consider a matrix A
along with a vector angles
that describes the target polynomial transformation. The qml.qsvt
function creates a corresponding circuit:
dev = qml.device("default.qubit", wires=2)
A = np.array([[0.1, 0.2], [0.3, 0.4]])
angles = np.array([0.1, 0.2, 0.3])
@qml.qnode(dev)
def example_circuit(A):
qml.qsvt(A, angles, wires=[0, 1])
return qml.expval(qml.PauliZ(wires=0))
This circuit is composed of qml.BlockEncode
and qml.PCPhase
operations.
>>> example_circuit(A)
tensor(0.97777078, requires_grad=True)
>>> print(example_circuit.qtape.expand(depth=1).draw(decimals=2))
0: ─╭∏_ϕ(0.30)─╭BlockEncode(M0)─╭∏_ϕ(0.20)─╭BlockEncode(M0)†─╭∏_ϕ(0.10)─┤ <Z>
1: ─╰∏_ϕ(0.30)─╰BlockEncode(M0)─╰∏_ϕ(0.20)─╰BlockEncode(M0)†─╰∏_ϕ(0.10)─┤
The qml.qsvt function creates a circuit that is targeted at simulators due to the use of matrix-based operations. For advanced users, you can use the operation-based qml.QSVT
template to perform the transformation with a custom choice of unitary and projector operations, which may be hardware compatible if a decomposition is provided.
The QSVT is a complex but powerful transformation capable of generalizing important algorithms like amplitude amplification. Stay tuned for a demo in the coming few weeks to learn more!
An updated QNode return system has been introduced. PennyLane QNodes now return exactly what you tell them to! 🎉 (#3957) (#3969) (#3946) (#3913) (#3914) (#3934)
This was an experimental feature introduced in version 0.25 of PennyLane that was enabled via qml.enable_return()
. Now, it's the default return system. Let's see how it works.
Consider the following circuit:
import pennylane as qml
dev = qml.device("default.qubit", wires=1)
@qml.qnode(dev)
def circuit(x):
qml.RX(x, wires=0)
return qml.expval(qml.PauliZ(0)), qml.probs(0)
In version 0.29 and earlier of PennyLane, circuit()
would return a single length-3 array:
>>> circuit(0.5)
tensor([0.87758256, 0.93879128, 0.06120872], requires_grad=True)
In versions 0.30 and above, circuit()
returns a length-2 tuple containing the expectation value and probabilities separately:
>>> circuit(0.5)
(tensor(0.87758256, requires_grad=True),
tensor([0.93879128, 0.06120872], requires_grad=True))
You can find more details about this change, along with help and troubleshooting tips to solve any issues. If you still have questions, comments, or concerns, we encourage you to post on the PennyLane discussion forum.
Single-qubit operations that have multi-qubit control can now be decomposed more efficiently using fewer CNOT gates. (#3851)
Three decompositions from arXiv:2302.06377 are provided and compare favourably to the already-available qml.ops.ctrl_decomp_zyz
:
wires = [0, 1, 2, 3, 4, 5]
control_wires = wires[1:]
@qml.qnode(qml.device('default.qubit', wires=6))
def circuit():
with qml.QueuingManager.stop_recording():
# the decomposition does not un-queue the target
target = qml.RX(np.pi/2, wires=0)
qml.ops.ctrl_decomp_bisect(target, (1,2,3,4,5))
return qml.state()
print(qml.draw(circuit, expansion_strategy="device")())
0: ──H─╭X──U(M0)─╭X──U(M0)†─╭X──U(M0)─╭X──U(M0)†──H─┤ State
1: ────├●────────│──────────├●────────│─────────────┤ State
2: ────├●────────│──────────├●────────│─────────────┤ State
3: ────╰●────────│──────────╰●────────│─────────────┤ State
4: ──────────────├●───────────────────├●────────────┤ State
5: ──────────────╰●───────────────────╰●────────────┤ State
A new decomposition to qml.SingleExcitation
has been added that halves the number of CNOTs required. (3976)
>>> qml.SingleExcitation.compute_decomposition(1.23, wires=(0,1))
[Adjoint(T(wires=[0])), Hadamard(wires=[0]), S(wires=[0]),
Adjoint(T(wires=[1])), Adjoint(S(wires=[1])), Hadamard(wires=[1]),
CNOT(wires=[1, 0]), RZ(-0.615, wires=[0]), RY(0.615, wires=[1]),
CNOT(wires=[1, 0]), Adjoint(S(wires=[0])), Hadamard(wires=[0]),
T(wires=[0]), Hadamard(wires=[1]), S(wires=[1]), T(wires=[1])]
The adjoint differentiation method can now be more efficient, avoiding the decomposition of operations that can be differentiated directly. Any operation that defines a generator()
can be differentiated with the adjoint method. (#3874)
For example, in version 0.29 the qml.CRY
operation would be decomposed when calculating the adjoint-method gradient. Executing the code below shows that this decomposition no longer takes place in version 0.30 and qml.CRY
is differentiated directly:
import jax
from jax import numpy as jnp
def compute_decomposition(self, phi, wires):
print("A decomposition has been performed!")
decomp_ops = [
qml.RY(phi / 2, wires=wires[1]),
qml.CNOT(wires=wires),
qml.RY(-phi / 2, wires=wires[1]),
qml.CNOT(wires=wires),
]
return decomp_ops
qml.CRY.compute_decomposition = compute_decomposition
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev, diff_method="adjoint")
def circuit(phi):
qml.Hadamard(wires=0)
qml.CRY(phi, wires=[0, 1])
return qml.expval(qml.PauliZ(1))
phi = jnp.array(0.5)
jax.grad(circuit)(phi)
Derivatives are computed more efficiently when using jax.jit
with gradient transforms; the trainable parameters are now set correctly instead of every parameter having to be set as trainable.
(#3697)
In the circuit below, only the derivative with respect to parameter b
is now calculated:
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev, interface="jax-jit")
def circuit(a, b):
qml.RX(a, wires=0)
qml.RY(b, wires=0)
qml.CNOT(wires=[0, 1])
return qml.expval(qml.PauliZ(0))
a = jnp.array(0.4)
b = jnp.array(0.5)
jac = jax.jacobian(circuit, argnums=[1])
jac_jit = jax.jit(jac)
jac_jit(a, b)
assert len(circuit.tape.trainable_params) == 1
In this release and future releases, we will be making changes to our device API with the goal in mind to make
developing plugins much easier for developers and unlock new device capabilities. Users shouldn't yet feel any of
these changes when using PennyLane, but here is what has changed this release:
Several functions in devices/qubit
have been added or improved:
sample_state
: returns a series of samples based on a given state vector and a number of shots. (#3720)
simulate
: supports measuring expectation values of large observables such as qml.Hamiltonian
, qml.SparseHamiltonian
, and qml.Sum
. (#3759)
apply_operation
: supports broadcasting. (#3852)
adjoint_jacobian
: supports adjoint differentiation in the new qubit state-vector device. (#3790)
qml.devices.qubit.preprocess
now allows circuits with non-commuting observables. (#3857)
qml.devices.qubit.measure
now computes the expectation values of Hamiltonian
and Sum
in a backpropagation-compatible way. (#3862)
Here are the functions, classes, and more that were added or improved to facilitate simulating ensembles of Rydberg atoms: (#3749) (#3911) (#3930) (#3936) (#3966) (#3987) (#3889) (#4021)
HardwareHamiltonian
: an internal class that contains additional information about pulses and settings.rydberg_interaction
: a user-facing function that returns a HardwareHamiltonian
containing the Hamiltonian of the interaction of all the Rydberg atoms.transmon_interaction
: a user-facing function for constructing the Hamiltonian that describes the circuit QED interaction Hamiltonian of superconducting transmon systems.drive
: a user-facing function function that returns a ParametrizedHamiltonian
(HardwareHamiltonian
) containing the Hamiltonian of the interaction between a driving electro-magnetic field and a group of qubits.rydberg_drive
: a user-facing function that returns a ParametrizedHamiltonian
(HardwareHamiltonian
) containing the Hamiltonian of the interaction between a driving laser field and a group of Rydberg atoms.max_distance
: a keyword argument added to qml.pulse.rydberg_interaction
to allow for the removal of negligible contributions from atoms beyond max_distance
from each other.ParametrizedEvolution
now takes two new Boolean keyword arguments: return_intermediate
and complementary
. They allow computing intermediate time evolution matrices. (#3900)
Activating return_intermediate
will return intermediate time evolution steps, for example for the matrix of the Operation, or of a quantum circuit when used in a QNode. Activating complementary
will make these intermediate steps be the remaining time evolution complementary to the output for complementary=False
. See the docstring for details.
Hardware-compatible pulse sequence gradients with qml.gradient.stoch_pulse_grad
can now be calculated faster using the new keyword argument use_broadcasting
. Executing a ParametrizedEvolution
that returns intermediate evolutions has increased performance using the state vector ODE solver, as well. (#4000) (#4004)
The QNode keyword argument mode
has been replaced by the boolean grad_on_execution
. (#3969)
The "default.gaussian"
device and parameter-shift CV both support the new return system, but only for single measurements. (#3946)
Keras and Torch NN modules are now compatible with the new return type system. (#3913) (#3914)
DefaultQutrit
now supports the new return system. (#3934)
The efficiency of tapering()
, tapering_hf()
and clifford()
have been improved. (3942)
The peak memory requirements of tapering()
and tapering_hf()
have been improved when used for larger observables. (3977)
Pauli arithmetic has been updated to convert to a Hamiltonian more efficiently. (#3939)
Operator
has a new Boolean attribute has_generator
. It returns whether or not the Operator
has a generator
defined. has_generator
is used in qml.operation.has_gen
, which improves its performance and extends differentiation support. (#3875)
The performance of CompositeOp
has been significantly improved now that it overrides determining whether it is being used with a batch of parameters (see Operator._check_batching
). Hamiltonian
also now overrides this, but it does nothing since it does not support batching. (#3915)
The performance of a Sum
operator has been significantly improved now that is_hermitian
checks that all coefficients are real if the operator has a pre-computed Pauli representation. (#3915)
The coefficients
function and the visualize
submodule of the qml.fourier
module now allow assigning different degrees for different parameters of the input function. (#3005)
Previously, the arguments degree
and filter_threshold
to qml.fourier.coefficients
were expected to be integers. Now, they can be a sequences of integers with one integer per function parameter (i.e. len(degree)==n_inputs
), resulting in a returned array with shape (2*degrees[0]+1,..., 2*degrees[-1]+1)
. The functions in qml.fourier.visualize
accordingly accept such arrays of coefficients.
A Shots
class has been added to the measurements
module to hold shot-related data. (#3682)
The custom JVP rules in PennyLane also now support non-scalar and mixed-shape tape parameters as well as multi-dimensional tape return types, like broadcasted qml.probs
, for example. (#3766)
The qchem.jordan_wigner
function has been extended to support more fermionic operator orders. (#3754) (#3751)
The AdaptiveOptimizer
has been updated to use non-default user-defined QNode arguments. (#3765)
Operators now use TensorLike
types dunder methods. (#3749)
qml.QubitStateVector.state_vector
now supports broadcasting. (#3852)
qml.SparseHamiltonian
can now be applied to any wires in a circuit rather than being restricted to all wires in the circuit. (#3888)
Operators can now be divided by scalars with /
with the addition of the Operation.__truediv__
dunder method. (#3749)
Printing an instance of MutualInfoMP
now displays the distribution of the wires between the two subsystems. (#3898)
Operator.num_wires
has been changed from an abstract value to AnyWires
. (#3919)
qml.transforms.sum_expand
is not run in Device.batch_transform
if the device supports Sum
observables. (#3915)
The type of n_electrons
in qml.qchem.Molecule
has been set to int
. (#3885)
Explicit errors have been added to QutritDevice
if classical_shadow
or shadow_expval
is measured. (#3934)
QubitDevice
now defines the private _get_diagonalizing_gates(circuit)
method and uses it when executing circuits. This allows devices that inherit from QubitDevice
to override and customize their definition of diagonalizing gates. (#3938)
retworkx
has been renamed to rustworkx
to accommodate the change in the package name. (#3975)
Exp
, Sum
, Prod
, and SProd
operator data is now a flat list instead of nested. (#3958) (#3983)
qml.transforms.convert_to_numpy_parameters
has been added to convert a circuit with interface-specific parameters to one with only numpy parameters. This transform is designed to replace qml.tape.Unwrap
. (#3899)
qml.operation.WiresEnum.AllWires
is now -2 instead of 0 to avoid the ambiguity between op.num_wires = 0
and op.num_wires = AllWires
. (#3978)
Execution code has been updated to use the new qml.transforms.convert_to_numpy_parameters
instead of qml.tape.Unwrap
. (#3989)
A sub-routine of expand_tape
has been converted into qml.tape.tape.rotations_and_diagonal_measurements
, a helper function that computes rotations and diagonal measurements for a tape with measurements with overlapping wires. (#3912)
Various operators and templates have been updated to ensure that their decompositions only return lists of operators. (#3243)
The qml.operation.enable_new_opmath
toggle has been introduced to cause dunder methods to return arithmetic operators instead of a Hamiltonian
or Tensor
. (#4008)
>>> type(qml.PauliX(0) @ qml.PauliZ(1))
<class 'pennylane.operation.Tensor'>
>>> qml.operation.enable_new_opmath()
>>> type(qml.PauliX(0) @ qml.PauliZ(1))
<class 'pennylane.ops.op_math.prod.Prod'>
>>> qml.operation.disable_new_opmath()
>>> type(qml.PauliX(0) @ qml.PauliZ(1))
<class 'pennylane.operation.Tensor'>
A new data class called Resources
has been added to store resources like the number of gates and circuit depth throughout a quantum circuit. (#3981)
A new function called _count_resources()
has been added to count the resources required when executing a QuantumTape
for a given number of shots. (#3996)
QuantumScript.specs
has been modified to make use of the new Resources
class. This also modifies the output of qml.specs()
. (#4015)
A new class called ResourcesOperation
has been added to allow users to define operations with custom resource information. (#4026)
For example, users can define a custom operation by inheriting from this new class:
>>> class CustomOp(qml.resource.ResourcesOperation):
... def resources(self):
... return qml.resource.Resources(num_wires=1, num_gates=2,
... gate_types={"PauliX": 2})
...
>>> CustomOp(wires=1)
CustomOp(wires=[1])
Then, we can track and display the resources of the workflow using qml.specs()
:
>>> dev = qml.device("default.qubit", wires=[0,1])
>>> @qml.qnode(dev)
... def circ():
... qml.PauliZ(wires=0)
... CustomOp(wires=1)
... return qml.state()
...
>>> print(qml.specs(circ)()['resources'])
wires: 2
gates: 3
depth: 1
shots: 0
gate_types:
{'PauliZ': 1, 'PauliX': 2}
MeasurementProcess.shape
now accepts a Shots
object as one of its arguments to reduce exposure to unnecessary execution details. (#4012)
The seed_recipes
argument has been removed from qml.classical_shadow
and qml.shadow_expval
. (#4020)
The tape method get_operation
has an updated signature. (#3998)
Both JIT interfaces are no longer compatible with JAX >0.4.3
(we raise an error for those versions). (#3877)
An operation that implements a custom generator
method, but does not always return a valid generator, also has to implement a has_generator
property that reflects in which scenarios a generator will be returned. (#3875)
Trainable parameters for the Jax interface are the parameters that are JVPTracer
, defined by setting argnums
. Previously, all JAX tracers, including those used for JIT compilation, were interpreted to be trainable. (#3697)
The keyword argument argnums
is now used for gradient transforms using Jax instead of argnum
. argnum
is automatically converted to argnums
when using Jax and will no longer be supported in v0.31 of PennyLane. (#3697) (#3847)
qml.OrbitalRotation
and, consequently, qml.GateFabric
are now more consistent with the interleaved Jordan-Wigner ordering. Previously, they were consistent with the sequential Jordan-Wigner ordering. (#3861)
Some MeasurementProcess
classes can now only be instantiated with arguments that they will actually use. For example, you can no longer create StateMP(qml.PauliX(0))
or PurityMP(eigvals=(-1,1), wires=Wires(0))
. (#3898)
Exp
, Sum
, Prod
, and SProd
operator data is now a flat list, instead of nested. (#3958) (#3983)
qml.tape.tape.expand_tape
and, consequentially, QuantumScript.expand
no longer update the input tape with rotations and diagonal measurements. Note that the newly expanded tape that is returned will still have the rotations and diagonal measurements. (#3912)
qml.Evolution
now initializes the coefficient with a factor of -1j
instead of 1j
. (#4024)
Nothing for this release!
The documentation of QubitUnitary
and DiagonalQubitUnitary
was clarified regarding the parameters of the operations. (#4031)
A typo has been corrected in the documentation for the introduction to inspecting_circuits
and chemistry
. (#3844)
Usage Details
and Theory
sections have been separated in the documentation for qml.qchem.taper_operation
. (3977)
ctrl_decomp_bisect
and ctrl_decomp_zyz
are no longer used by default when decomposing controlled operations due to the presence of a global phase difference in the zyz decomposition of some target operators.
Fixed a bug where qml.math.dot
returned a numpy array instead of an autograd array, breaking autograd derivatives in certain circumstances. (#4019)
Operators now cast a tuple
to an np.ndarray
as well as list
. (#4022)
Fixed a bug where qml.ctrl
with parametric gates was incompatible with PyTorch tensors on GPUs. (#4002)
Fixed a bug where the broadcast expand results were stacked along the wrong axis for the new return system. (#3984)
A more informative error message is raised in qml.jacobian
to explain potential problems with the new return types specification. (#3997)
Fixed a bug where calling Evolution.generator
with coeff
being a complex ArrayBox raised an error. (#3796)
MeasurementProcess.hash
now uses the hash property of the observable. The property now depends on all properties that affect the behaviour of the object, such as VnEntropyMP.log_base
or the distribution of wires between the two subsystems in MutualInfoMP
. (#3898)
The enum measurements.Purity
has been added so that PurityMP.return_type
is defined. str
and repr
for PurityMP
are also now defined. (#3898)
Sum.hash
and Prod.hash
have been changed slightly to work with non-numeric wire labels. sum_expand
should now return correct results and not treat some products as the same operation. (#3898)
Fixed bug where the coefficients where not ordered correctly when summing a ParametrizedHamiltonian
with other operators. (#3749) (#3902)
The metric tensor transform is now fully compatible with Jax and therefore users can provide multiple parameters. (#3847)
qml.math.ndim
and qml.math.shape
are now registered for built-ins and autograd to accomodate Autoray 0.6.1. #3864
Ensured that qml.data.load
returns datasets in a stable and expected order. (#3856)
The qml.equal
function now handles comparisons of ParametrizedEvolution
operators. (#3870)
qml.devices.qubit.apply_operation
catches the tf.errors.UnimplementedError
that occurs when PauliZ
or CNOT
gates are applied to a large (>8 wires) tensorflow state. When that occurs, the logic falls back to the tensordot logic instead. (#3884)
Fixed parameter broadcasting support with qml.counts
in most cases and introduced explicit errors otherwise. (#3876)
An error is now raised if a QNode with Jax-jit in use returns counts
while having trainable parameters (#3892)
A correction has been added to the reference values in test_dipole_of
to account for small changes (~2e-8
) in the computed dipole moment values resulting from the new PySCF 2.2.0 release. (#3908)
SampleMP.shape
is now correct when sampling only occurs on a subset of the device wires. (#3921)
An issue has been fixed in qchem.Molecule
to allow basis sets other than the hard-coded ones to be used in the Molecule
class. (#3955)
Fixed bug where all devices that inherit from DefaultQubit
claimed to support ParametrizedEvolution
. Now, only DefaultQubitJax
supports the operator, as expected. (#3964)
Ensured that parallel AnnotatedQueues
do not queue each other's contents. (#3924)
Added a map_wires
method to PauliWord
and PauliSentence
, and ensured that operators call it in their respective map_wires
methods if they have a Pauli rep. (#3985)
Fixed a bug when a Tensor
is multiplied by a Hamiltonian
or vice versa. (#4036)
This release contains contributions from (in alphabetical order):
Komi Amiko,
Utkarsh Azad,
Thomas Bromley,
Isaac De Vlugt,
Olivia Di Matteo,
Lillian M. A. Frederiksen,
Diego Guala,
Soran Jahangiri,
Korbinian Kottmann,
Christina Lee,
Vincent Michaud-Rioux,
Albert Mitjans Coma,
Romain Moyard,
Lee J. O'Riordan,
Mudit Pandey,
Matthew Silverman,
Jay Soni,
David Wierichs.
Published by albi3ro over 1 year ago
qml.math.ndim
and qml.math.shape
for builtins and autograd. Accommodates changes made by Autograd v0.6.1.This release contains contributions from (in alphabetical order):
Christina Lee
Published by albi3ro over 1 year ago
Support for creating pulse-based circuits that describe evolution under a time-dependent Hamiltonian has now been added, as well as the ability to execute and differentiate these pulse-based circuits on simulator.
(#3586)(#3617)(#3645)(#3652)(#3665)(#3673)(#3706)(#3730)
A time-dependent Hamiltonian can be created using qml.pulse.ParametrizedHamiltonian
, which holds information representing a linear combination of operators with parametrized coefficents and can be constructed as follows:
from jax import numpy as jnp
f1 = lambda p, t: p * jnp.sin(t) * (t - 1)
f2 = lambda p, t: p[0] * jnp.cos(p[1]* t ** 2)
XX = qml.PauliX(0) @ qml.PauliX(1)
YY = qml.PauliY(0) @ qml.PauliY(1)
ZZ = qml.PauliZ(0) @ qml.PauliZ(1)
H = 2 * ZZ + f1 * XX + f2 * YY
>>> H
ParametrizedHamiltonian: terms=3
>>> p1 = jnp.array(1.2)
>>> p2 = jnp.array([2.3, 3.4])
>>> H((p1, p2), t=0.5)
(2*(PauliZ(wires=[0]) @ PauliZ(wires=[1]))) + ((-0.2876553231625218*(PauliX(wires=[0]) @ PauliX(wires=[1]))) + (1.517961235535459*(PauliY(wires=[0]) @ PauliY(wires=[1]))))
The time-dependent Hamiltonian can be used within a circuit with qml.evolve
:
def pulse_circuit(params, time):
qml.evolve(H)(params, time)
return qml.expval(qml.PauliX(0) @ qml.PauliY(1))
Pulse-based circuits can be executed and differentiated on the default.qubit.jax
simulator using JAX as an interface:
>>> dev = qml.device("default.qubit.jax", wires=2)
>>> qnode = qml.QNode(pulse_circuit, dev, interface="jax")
>>> params = (p1, p2)
>>> qnode(params, time=0.5)
Array(0.72153819, dtype=float64)
>>> jax.grad(qnode)(params, time=0.5)
(Array(-0.11324919, dtype=float64),
Array([-0.64399616, 0.06326374], dtype=float64))
Check out the qml.pulse documentation page for more details!
A new operation qml.SpecialUnitary
has been added, providing access to an arbitrary unitary gate via a parametrization in the Pauli basis.
(#3650) (#3651) (#3674)
qml.SpecialUnitary
creates a unitary that exponentiates a linear combination of all possible Pauli words in lexicographical order — except for the identity operator — for num_wires
wires, of which there are 4**num_wires - 1
. As its first argument, qml.SpecialUnitary
takes a list of the 4**num_wires - 1
parameters that are the coefficients of the linear combination.
To see all possible Pauli words for num_wires
wires, you can use the qml.ops.qubit.special_unitary.pauli_basis_strings
function:
>>> qml.ops.qubit.special_unitary.pauli_basis_strings(1) # 4**1-1 = 3 Pauli words
['X', 'Y', 'Z']
>>> qml.ops.qubit.special_unitary.pauli_basis_strings(2) # 4**2-1 = 15 Pauli words
['IX', 'IY', 'IZ', 'XI', 'XX', 'XY', 'XZ', 'YI', 'YX', 'YY', 'YZ', 'ZI', 'ZX', 'ZY', 'ZZ']
To use qml.SpecialUnitary
, for example, on a single qubit, we may define
>>> thetas = np.array([0.2, 0.1, -0.5])
>>> U = qml.SpecialUnitary(thetas, 0)
>>> qml.matrix(U)
array([[ 0.8537127 -0.47537233j, 0.09507447+0.19014893j],
[-0.09507447+0.19014893j, 0.8537127 +0.47537233j]])
A single non-zero entry in the parameters will create a Pauli rotation:
>>> x = 0.412
>>> theta = x * np.array([1, 0, 0]) # The first entry belongs to the Pauli word "X"
>>> su = qml.SpecialUnitary(theta, wires=0)
>>> rx = qml.RX(-2 * x, 0) # RX introduces a prefactor -0.5 that has to be compensated
>>> qml.math.allclose(qml.matrix(su), qml.matrix(rx))
True
This operation can be differentiated with hardware-compatible methods like parameter shifts and it supports parameter broadcasting/batching, but not both at the same time. Learn more by visiting the qml.SpecialUnitary documentation.
The Hadamard test gradient transform is now available via qml.gradients.hadamard_grad
. This transform is also available as a differentiation method within QNode
s. (#3625) (#3736)
qml.gradients.hadamard_grad
is a hardware-compatible transform that calculates the gradient of a quantum circuit using the Hadamard test. Note that the device requires an auxiliary wire to calculate the gradient.
>>> dev = qml.device("default.qubit", wires=2)
>>> @qml.qnode(dev)
... def circuit(params):
... qml.RX(params[0], wires=0)
... qml.RY(params[1], wires=0)
... qml.RX(params[2], wires=0)
... return qml.expval(qml.PauliZ(0))
>>> params = np.array([0.1, 0.2, 0.3], requires_grad=True)
>>> qml.gradients.hadamard_grad(circuit)(params)
(tensor(-0.3875172, requires_grad=True),
tensor(-0.18884787, requires_grad=True),
tensor(-0.38355704, requires_grad=True))
This transform can be registered directly as the quantum gradient transform to use during autodifferentiation:
>>> dev = qml.device("default.qubit", wires=2)
>>> @qml.qnode(dev, interface="jax", diff_method="hadamard")
... def circuit(params):
... qml.RX(params[0], wires=0)
... qml.RY(params[1], wires=0)
... qml.RX(params[2], wires=0)
... return qml.expval(qml.PauliZ(0))
>>> params = jax.numpy.array([0.1, 0.2, 0.3])
>>> jax.jacobian(circuit)(params)
Array([-0.3875172 , -0.18884787, -0.38355705], dtype=float32)
The gradient transform qml.gradients.spsa_grad
is now registered as a differentiation method for QNodes.
(#3440)
The SPSA gradient transform can now be used implicitly by marking a QNode as differentiable with SPSA. It can be selected via
>>> dev = qml.device("default.qubit", wires=1)
>>> @qml.qnode(dev, interface="jax", diff_method="spsa", h=0.05, num_directions=20)
... def circuit(x):
... qml.RX(x, 0)
... return qml.expval(qml.PauliZ(0))
>>> jax.jacobian(circuit)(jax.numpy.array(0.5))
Array(-0.4792258, dtype=float32, weak_type=True)
The argument num_directions
determines how many directions of simultaneous perturbation are used and therefore the number of circuit evaluations, up to a prefactor. See the SPSA gradient transform documentation for details. Note: The full SPSA optimization method is already available as qml.SPSAOptimizer
.
The default interface is now auto
. There is no need to specify the interface anymore; it is automatically determined by checking your QNode parameters.
(#3677)(#3752) (#3829)
import jax
import jax.numpy as jnp
qml.enable_return()
a = jnp.array(0.1)
b = jnp.array(0.2)
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def circuit(a, b):
qml.RY(a, wires=0)
qml.RX(b, wires=1)
qml.CNOT(wires=[0, 1])
return qml.expval(qml.PauliZ(0)), qml.expval(qml.PauliY(1))
>>> circuit(a, b)
(Array(0.9950042, dtype=float32), Array(-0.19767681, dtype=float32))
>>> jac = jax.jacobian(circuit)(a, b)
>>> jac
(Array(-0.09983341, dtype=float32, weak_type=True), Array(0.01983384, dtype=float32, weak_type=True))
The JAX-JIT interface now supports higher-order gradient computation with the new return types system.
(#3498)
Here is an example of using JAX-JIT to compute the Hessian of a circuit:
import pennylane as qml
import jax
from jax import numpy as jnp
jax.config.update("jax_enable_x64", True)
qml.enable_return()
dev = qml.device("default.qubit", wires=2)
@jax.jit
@qml.qnode(dev, interface="jax-jit", diff_method="parameter-shift", max_diff=2)
def circuit(a, b):
qml.RY(a, wires=0)
qml.RX(b, wires=1)
return qml.expval(qml.PauliZ(0)), qml.expval(qml.PauliZ(1))
a, b = jnp.array(1.0), jnp.array(2.0)
>>> jax.hessian(circuit, argnums=[0, 1])(a, b)
(((Array(-0.54030231, dtype=float64, weak_type=True),
Array(0., dtype=float64, weak_type=True)),
(Array(-1.76002563e-17, dtype=float64, weak_type=True),
Array(0., dtype=float64, weak_type=True))),
((Array(0., dtype=float64, weak_type=True),
Array(-1.00700085e-17, dtype=float64, weak_type=True)),
(Array(0., dtype=float64, weak_type=True),
Array(0.41614684, dtype=float64, weak_type=True))))
The qchem
workflow has been modified to support both Autograd and JAX frameworks.
(#3458) (#3462) (#3495)
The JAX interface is automatically used when the differentiable parameters are JAX objects. Here is an example for computing the Hartree-Fock energy gradients with respect to the atomic coordinates.
import pennylane as qml
from pennylane import numpy as np
import jax
symbols = ["H", "H"]
geometry = np.array([[0.0, 0.0, 0.0], [0.0, 0.0, 1.0]])
mol = qml.qchem.Molecule(symbols, geometry)
args = [jax.numpy.array(mol.coordinates)]
>>> jax.grad(qml.qchem.hf_energy(mol))(*args)
Array([[ 0. , 0. , 0.3650435],
[ 0. , 0. , -0.3650435]], dtype=float64)
The kernel matrix utility functions in qml.kernels
are now autodifferentiation-compatible. In addition, they support batching, for example for quantum kernel execution with shot vectors.
(#3742)
This allows for the following:
dev = qml.device('default.qubit', wires=2, shots=(100, 100))
@qml.qnode(dev)
def circuit(x1, x2):
qml.templates.AngleEmbedding(x1, wires=dev.wires)
qml.adjoint(qml.templates.AngleEmbedding)(x2, wires=dev.wires)
return qml.probs(wires=dev.wires)
kernel = lambda x1, x2: circuit(x1, x2)
We can then compute the kernel matrix on a set of 4 (random) feature vectors X
but using two sets of 100 shots each via
>>> X = np.random.random((4, 2))
>>> qml.kernels.square_kernel_matrix(X, kernel)[:, 0]
tensor([[[1. , 0.86, 0.88, 0.92],
[0.86, 1. , 0.75, 0.97],
[0.88, 0.75, 1. , 0.91],
[0.92, 0.97, 0.91, 1. ]],
[[1. , 0.93, 0.91, 0.92],
[0.93, 1. , 0.8 , 1. ],
[0.91, 0.8 , 1. , 0.91],
[0.92, 1. , 0.91, 1. ]]], requires_grad=True)
Note that we have extracted the first probability vector entry for each 100-shot evaluation.
Hamiltonian evolution using qml.evolve
or qml.exp
can now be decomposed into operations.
(#3691) (#3777)
If the time-evolved Hamiltonian is equivalent to another PennyLane operation, then that operation is returned as the decomposition:
>>> exp_op = qml.evolve(qml.PauliX(0) @ qml.PauliX(1))
>>> exp_op.decomposition()
[IsingXX((2+0j), wires=[0, 1])]
If the Hamiltonian is a Pauli word, then the decomposition is provided as a qml.PauliRot
operation:
>>> qml.evolve(qml.PauliZ(0) @ qml.PauliX(1)).decomposition()
[PauliRot((2+0j), ZX, wires=[0, 1])]
Otherwise, the Hamiltonian is a linear combination of operators and the Suzuki-Trotter decomposition is used:
>>> qml.evolve(qml.sum(qml.PauliX(0), qml.PauliY(0), qml.PauliZ(0)), num_steps=2).decomposition()
[RX((1+0j), wires=[0]),
RY((1+0j), wires=[0]),
RZ((1+0j), wires=[0]),
RX((1+0j), wires=[0]),
RY((1+0j), wires=[0]),
RZ((1+0j), wires=[0])]
A new method called qml.qchem.givens_decomposition
has been added, which decomposes a unitary into a sequence of Givens rotation gates with phase shifts and a diagonal phase matrix.
(#3573)
unitary = np.array([[ 0.73678+0.27511j, -0.5095 +0.10704j, -0.06847+0.32515j],
[-0.21271+0.34938j, -0.38853+0.36497j, 0.61467-0.41317j],
[ 0.41356-0.20765j, -0.00651-0.66689j, 0.32839-0.48293j]])
phase_mat, ordered_rotations = qml.qchem.givens_decomposition(unitary)
>>> phase_mat
tensor([-0.20604358+0.9785369j , -0.82993272+0.55786114j,
0.56230612-0.82692833j], requires_grad=True)
>>> ordered_rotations
[(tensor([[-0.65087861-0.63937521j, -0.40933651-0.j ],
[-0.29201359-0.28685265j, 0.91238348-0.j ]], requires_grad=True),
(0, 1)),
(tensor([[ 0.47970366-0.33308926j, -0.8117487 -0.j ],
[ 0.66677093-0.46298215j, 0.5840069 -0.j ]], requires_grad=True),
(1, 2)),
(tensor([[ 0.36147547+0.73779454j, -0.57008306-0.j ],
[ 0.2508207 +0.51194108j, 0.82158706-0.j ]], requires_grad=True),
(0, 1))]
A new template called qml.BasisRotation
has been added, which performs a basis transformation defined by a set of fermionic ladder operators.
(#3573)
import pennylane as qml
from pennylane import numpy as np
V = np.array([[ 0.53672126+0.j , -0.1126064 -2.41479668j],
[-0.1126064 +2.41479668j, 1.48694623+0.j ]])
eigen_vals, eigen_vecs = np.linalg.eigh(V)
umat = eigen_vecs.T
wires = range(len(umat))
def circuit():
qml.adjoint(qml.BasisRotation(wires=wires, unitary_matrix=umat))
for idx, eigenval in enumerate(eigen_vals):
qml.RZ(eigenval, wires=[idx])
qml.BasisRotation(wires=wires, unitary_matrix=umat)
>>> circ_unitary = qml.matrix(circuit)()
>>> np.round(circ_unitary/circ_unitary[0][0], 3)
tensor([[ 1. -0.j , -0. +0.j , -0. +0.j , -0. +0.j ],
[-0. +0.j , -0.516-0.596j, -0.302-0.536j, -0. +0.j ],
[-0. +0.j , 0.35 +0.506j, -0.311-0.724j, -0. +0.j ],
[-0. +0.j , -0. +0.j , -0. +0.j , -0.438+0.899j]], requires_grad=True)
A new function called qml.qchem.load_basisset
has been added to extract qml.qchem
basis set data from the Basis Set Exchange library.
(#3363)
A new function called qml.math.max_entropy
has been added to compute the maximum entropy of a quantum state.
(#3594)
A new template called qml.TwoLocalSwapNetwork
has been added that implements a canonical 2-complete linear (2-CCL) swap network described in arXiv:1905.05118.
(#3447)
dev = qml.device('default.qubit', wires=5)
weights = np.random.random(size=qml.templates.TwoLocalSwapNetwork.shape(len(dev.wires)))
acquaintances = lambda index, wires, param: (qml.CRY(param, wires=index)
if np.abs(wires[0]-wires[1]) else qml.CRZ(param, wires=index))
@qml.qnode(dev)
def swap_network_circuit():
qml.templates.TwoLocalSwapNetwork(dev.wires, acquaintances, weights, fermionic=False)
return qml.state()
>>> print(weights)
tensor([0.20308242, 0.91906199, 0.67988804, 0.81290256, 0.08708985,
0.81860084, 0.34448344, 0.05655892, 0.61781612, 0.51829044], requires_grad=True)
>>> print(qml.draw(swap_network_circuit, expansion_strategy = 'device')())
0: ─╭●────────╭SWAP─────────────────╭●────────╭SWAP─────────────────╭●────────╭SWAP─┤ State
1: ─╰RY(0.20)─╰SWAP─╭●────────╭SWAP─╰RY(0.09)─╰SWAP─╭●────────╭SWAP─╰RY(0.62)─╰SWAP─┤ State
2: ─╭●────────╭SWAP─╰RY(0.68)─╰SWAP─╭●────────╭SWAP─╰RY(0.34)─╰SWAP─╭●────────╭SWAP─┤ State
3: ─╰RY(0.92)─╰SWAP─╭●────────╭SWAP─╰RY(0.82)─╰SWAP─╭●────────╭SWAP─╰RY(0.52)─╰SWAP─┤ State
4: ─────────────────╰RY(0.81)─╰SWAP─────────────────╰RY(0.06)─╰SWAP─────────────────┤ State
A new function called qml.pulse.pwc
has been added as a convenience function for defining a qml.pulse.ParametrizedHamiltonian
. This function can be used to create a callable coefficient by setting the timespan over which the function should be non-zero. The resulting callable can be passed an array of parameters and a time.
(#3645)
>>> timespan = (2, 4)
>>> f = qml.pulse.pwc(timespan)
>>> f * qml.PauliX(0)
ParametrizedHamiltonian: terms=1
The params
array will be used as bin values evenly distributed over the timespan, and the parameter t
will determine which of the bins is returned.
>>> f(params=[1.2, 2.3, 3.4, 4.5], t=3.9)
DeviceArray(4.5, dtype=float32)
>>> f(params=[1.2, 2.3, 3.4, 4.5], t=6) # zero outside the range (2, 4)
DeviceArray(0., dtype=float32)
A new function calledqml.pulse.pwc_from_function
has been added as a decorator for defining a qml.pulse.ParametrizedHamiltonian
. This function can be used to decorate a function and create a piecewise constant approximation of it.
(#3645)
>>> @qml.pulse.pwc_from_function((2, 4), num_bins=10)
... def f1(p, t):
... return p * t
The resulting function approximates the same of p**2 * t
on the interval t=(2, 4)
in 10 bins, and returns zero outside the interval.
# t=2 and t=2.1 are within the same bin
>>> f1(3, 2), f1(3, 2.1)
(DeviceArray(6., dtype=float32), DeviceArray(6., dtype=float32))
# next bin
>>> f1(3, 2.2)
DeviceArray(6.6666665, dtype=float32)
# outside the interval t=(2, 4)
>>> f1(3, 5)
DeviceArray(0., dtype=float32)
Add ParametrizedHamiltonianPytree
class, which is a pytree jax object representing a parametrized Hamiltonian, where the matrix computation is delayed to improve performance.
(#3779)
The function qml.dot
has been updated to compute the dot product between a vector and a list of operators.
(#3586)
>>> coeffs = np.array([1.1, 2.2])
>>> ops = [qml.PauliX(0), qml.PauliY(0)]
>>> qml.dot(coeffs, ops)
(1.1*(PauliX(wires=[0]))) + (2.2*(PauliY(wires=[0])))
>>> qml.dot(coeffs, ops, pauli=True)
1.1 * X(0) + 2.2 * Y(0)
qml.evolve
returns the evolution of an Operator
or a ParametrizedHamiltonian
.
(#3617) (#3706)
qml.ControlledQubitUnitary
now inherits from qml.ops.op_math.ControlledOp
, which defines decomposition
, expand
, and sparse_matrix
rather than raising an error.
(#3450)
Parameter broadcasting support has been added for the qml.ops.op_math.Controlled
class if the base operator supports broadcasting.
(#3450)
The qml.generator
function now checks if the generator is Hermitian, rather than whether it is a subclass of Observable
. This allows it to return valid generators from SymbolicOp
and CompositeOp
classes.
(#3485)
The qml.equal
function has been extended to compare Prod
and Sum
operators.
(#3516)
qml.purity
has been added as a measurement process for purity
(#3551)
In-place inversion has been removed for qutrit operations in preparation for the removal of in-place inversion.
(#3566)
The qml.utils.sparse_hamiltonian
function has been moved to thee qml.Hamiltonian.sparse_matrix
method.
(#3585)
The qml.pauli.PauliSentence.operation()
method has been improved to avoid instantiating an SProd
operator when the coefficient is equal to 1.
(#3595)
Batching is now allowed in all SymbolicOp
operators, which include Exp
, Pow
and SProd
.
(#3597)
The Sum
and Prod
operations now have broadcasted operands.
(#3611)
The XYX single-qubit unitary decomposition has been implemented.
(#3628)
All dunder methods now return NotImplemented
, allowing the right dunder method (e.g. __radd__
) of the other class to be called.
(#3631)
The qml.GellMann
operators now include their index when displayed.
(#3641)
qml.ops.ctrl_decomp_zyz
has been added to compute the decomposition of a controlled single-qubit operation given a single-qubit operation and the control wires.
(#3681)
qml.pauli.is_pauli_word
now supports Prod
and SProd
operators, and it returns False
when a Hamiltonian
contains more than one term.
(#3692)
qml.pauli.pauli_word_to_string
now supports Prod
, SProd
and Hamiltonian
operators.
(#3692)
qml.ops.op_math.Controlled
can now decompose single qubit target operations more effectively using the ZYZ decomposition.
(#3726)
The qml.qchem.Molecule
class raises an error when the molecule has an odd number of electrons or when the spin multiplicity is not 1.
(#3748)
qml.qchem.basis_rotation
now accounts for spin, allowing it to perform Basis Rotation Groupings for molecular hamiltonians.
(#3714)(#3774)
The gradient transforms work for the new return type system with non-trivial classical jacobians.
(#3776)
The default.mixed
device has received a performance improvement for multi-qubit operations. This also allows to apply channels that act on more than seven qubits, which was not possible before.
(#3584)
qml.dot
now groups coefficients together.
(#3691)
>>> qml.dot(coeffs=[2, 2, 2], ops=[qml.PauliX(0), qml.PauliY(1), qml.PauliZ(2)])
2*(PauliX(wires=[0]) + PauliY(wires=[1]) + PauliZ(wires=[2]))
qml.generator
now supports operators with Sum
and Prod
generators.
(#3691)
The Sum._sort
method now takes into account the name of the operator when sorting.
(#3691)
A new tape transform called qml.transforms.sign_expand
has been added. It implements the optimal decomposition of a fast forwardable Hamiltonian that minimizes the variance of its estimator in the Single-Qubit-Measurement from arXiv:2207.09479.
(#2852)
The qml.math
module now also contains a submodule for fast Fourier transforms, qml.math.fft
.
(#1440)
The submodule in particular provides differentiable versions of the following functions, available in all common interfaces for PennyLane
Note that the output of the derivative of these functions may differ when used with complex-valued inputs, due to different conventions on complex-valued derivatives.
Validation has been added on gradient keyword arguments when initializing a QNode — if unexpected keyword arguments are passed, a UserWarning
is raised. A list of the current expected gradient function keyword arguments can be accessed via qml.gradients.SUPPORTED_GRADIENT_KWARGS
.
(#3526)
The numpy
version has been constrained to <1.24
.
(#3563)
Support for two-qubit unitary decomposition with JAX-JIT has been added.
(#3569)
qml.math.size
now supports PyTorch tensors.
(#3606)
Most quantum channels are now fully differentiable on all interfaces.
(#3612)
qml.math.matmul
now supports PyTorch and Autograd tensors.
(#3613)
Add qml.math.detach
, which detaches a tensor from its trace. This stops automatic gradient computations.
(#3674)
Add typing.TensorLike
type.
(#3675)
qml.QuantumMonteCarlo
template is now JAX-JIT compatible when passing jax.numpy
arrays to the template.
(#3734)
DefaultQubitJax
now supports evolving the state vector when executing qml.pulse.ParametrizedEvolution
gates.
(#3743)
SProd.sparse_matrix
now supports interface-specific variables with a single element as the scalar
.
(#3770)
Added argnum
argument to metric_tensor
. By passing a sequence of indices referring to trainable tape parameters, the metric tensor is only computed with respect to these parameters. This reduces the number of tapes that have to be run.
(#3587)
The parameter-shift derivative of variances saves a redundant evaluation of the corresponding unshifted expectation value tape, if possible
(#3744)
The apply_operation
single-dispatch function is added to devices/qubit
that applies an operation to a state and returns a new state.
(#3637)
The preprocess
function is added to devices/qubit
that validates, expands, and transforms a batch of QuantumTape
objects to abstract preprocessing details away from the device.
(#3708)
The create_initial_state
function is added to devices/qubit
that returns an initial state for an execution.
(#3683)
The simulate
function is added to devices/qubit
that turns a single quantum tape into a measurement result. The function only supports state based measurements with either no observables or observables with diagonalizing gates. It supports simultaneous measurement of non-commuting observables.
(#3700)
The ExecutionConfig
data class has been added.
(#3649)
The StatePrep
class has been added as an interface that state-prep operators must implement.
(#3654)
qml.QubitStateVector
now implements the StatePrep
interface.
(#3685)
qml.BasisState
now implements the StatePrep
interface.
(#3693)
New Abstract Base Class for devices Device
is added to the devices.experimental
submodule. This interface is still in experimental mode and not integrated with the rest of pennylane.
(#3602)
Writing Hamiltonians to a file using the qml.data
module has been improved by employing a condensed writing format.
(#3592)
Lazy-loading in the qml.data.Dataset.read()
method is more universally supported.
(#3605)
The qchem.Molecule
class raises an error when the molecule has an odd number of electrons or when the spin multiplicity is not 1.
(#3748)
qml.draw
and qml.draw_mpl
have been updated to draw any quantum function, which allows for visualizing only part of a complete circuit/QNode.
(#3760)
The string representation of a Measurement Process now includes the _eigvals
property if it is set.
(#3820)
The argument mode
in execution has been replaced by the boolean grad_on_execution
in the new execution pipeline.
(#3723)
qml.VQECost
has been removed.
(#3735)
The default interface is now auto
.
(#3677)(#3752)(#3829)
The interface is determined during the QNode call instead of the initialization. It means that the gradient_fn
and gradient_kwargs
are only defined on the QNode at the beginning of the call. Moreover, without specifying the interface it is not possible to guarantee that the device will not be changed during the call if you are using backprop (such as default.qubit
changing to default.qubit.jax
) whereas before it was happening at initialization.
The tape method get_operation
can also now return the operation index in the tape, and it can be activated by setting the return_op_index
to True
: get_operation(idx, return_op_index=True)
. It will become the default in version 0.30
.
(#3667)
Operation.inv()
and the Operation.inverse
setter have been removed. Please use qml.adjoint
or qml.pow
instead.
(#3618)
For example, instead of
>>> qml.PauliX(0).inv()
use
>>> qml.adjoint(qml.PauliX(0))
The Operation.inverse
property has been removed completely.
(#3725)
The target wires of qml.ControlledQubitUnitary
are no longer available via op.hyperparameters["u_wires"]
. Instead, they can be accesses via op.base.wires
or op.target_wires
.
(#3450)
The tape constructed by a QNode
is no longer queued to surrounding contexts.
(#3509)
Nested operators like Tensor
, Hamiltonian
, and Adjoint
now remove their owned operators from the queue instead of updating their metadata to have an "owner"
.
(#3282)
qml.qchem.scf
, qml.RandomLayers.compute_decomposition
, and qml.Wires.select_random
now use local random number generators instead of global random number generators. This may lead to slightly different random numbers and an independence of the results from the global random number generation state. Please provide a seed to each individual function instead if you want controllable results.
(#3624)
qml.transforms.measurement_grouping
has been removed. Users should use qml.transforms.hamiltonian_expand
instead.
(#3701)
op.simplify()
for operators which are linear combinations of Pauli words will use a builtin Pauli representation to more efficiently compute the simplification of the operator.
(#3481)
All Operator
's input parameters that are lists are cast into vanilla numpy arrays.
(#3659)
QubitDevice.expval
no longer permutes an observable's wire order before passing it to QubitDevice.probability
. The associated downstream changes for default.qubit
have been made, but this may still affect expectations for other devices that inherit from QubitDevice
and override probability
(or any other helper functions that take a wire order such as marginal_prob
, estimate_probability
or analytic_probability
).
(#3753)
qml.utils.sparse_hamiltonian
function has been deprecated, and usage will now raise a warning. Instead, one should use the qml.Hamiltonian.sparse_matrix
method.
(#3585)
qml.op_sum
has been deprecated. Users should use qml.sum
instead.
(#3686)
The use of Evolution
directly has been deprecated. Users should use qml.evolve
instead. This new function changes the sign of the given parameter.
(#3706)
Use of qml.dot
with a QNodeCollection
has been deprecated.
(#3586)
Revise note on GPU support in the circuit introduction.
(#3836)
Make warning about vanilla version of NumPy for differentiation more prominent.
(#3838)
The documentation for qml.operation
has been improved.
(#3664)
The code example in qml.SparseHamiltonian
has been updated with the correct wire range.
(#3643)
A hyperlink has been added in the text for a URL in the qml.qchem.mol_data
docstring.
(#3644)
A typo was corrected in the documentation for qml.math.vn_entropy
.
(#3740)
Fixed a bug where measuring qml.probs
in the computational basis with non-commuting measurements returned incorrect results. Now an error is raised.
(#3811)
Fixed a bug in the drawer where nested controlled operations would output the label of the operation being controlled, rather than the control values.
(#3745)
Fixed a bug in qml.transforms.metric_tensor
where prefactors of operation generators were taken into account multiple times, leading to wrong outputs for non-standard operations.
(#3579)
Local random number generators are now used where possible to avoid mutating the global random state.
(#3624)
The networkx
version change being broken has been fixed by selectively skipping a qcut
TensorFlow-JIT test.
(#3609)(#3619)
Fixed the wires for the Y
decomposition in the ZX calculus transform.
(#3598)
qml.pauli.PauliWord
is now pickle-able.
(#3588)
Child classes of QuantumScript
now return their own type when using SomeChildClass.from_queue
.
(#3501)
A typo has been fixed in the calculation and error messages in operation.py
(#3536)
qml.data.Dataset.write()
now ensures that any lazy-loaded values are loaded before they are written to a file.
(#3605)
Tensor._batch_size
is now set to None
during initialization, copying and map_wires
.
(#3642)(#3661)
Tensor.has_matrix
is now set to True
.
(#3647)
Fixed typo in the example of qml.IsingZZ
gate decomposition.
(#3676)
Fixed a bug that made tapes/qnodes using qml.Snapshot
incompatible with qml.drawer.tape_mpl
.
(#3704)
Tensor._pauli_rep
is set to None
during initialization and Tensor.data
has been added to its setter.
(#3722)
qml.math.ndim
has been redirected to jnp.ndim
when using it on a jax
tensor.
(#3730)
Implementations of marginal_prob
(and subsequently, qml.probs
) now return probabilities with the expected wire order.
(#3753)
This bug affected most probabilistic measurement processes on devices that inherit from QubitDevice
when the measured wires are out of order with respect to the device wires and 3 or more wires are measured. The assumption was that marginal probabilities would be computed with the device's state and wire order, then re-ordered according to the measurement process wire order. Instead, the re-ordering went in the inverse direction (that is, from measurement process wire order to device wire order). This is now fixed. Note that this only occurred for 3 or more measured wires because this mapping is identical otherwise. More details and discussion of this bug can be found in the original bug report.
Empty iterables can no longer be returned from QNodes.
(#3769)
The keyword arguments for qml.equal
now are used when comparing the observables of a Measurement Process. The eigvals of measurements are only requested if both observables are None
, saving computational effort.
(#3820)
Only converts input to qml.Hermitian
to a numpy array if the input is a list.
(#3820)
This release contains contributions from (in alphabetical order):
Gian-Luca Anselmetti,
Guillermo Alonso-Linaje,
Juan Miguel Arrazola,
Ikko Ashimine,
Utkarsh Azad,
Miriam Beddig,
Cristian Boghiu,
Thomas Bromley,
Astral Cai,
Isaac De Vlugt,
Olivia Di Matteo,
Lillian M. A. Frederiksen,
Soran Jahangiri,
Korbinian Kottmann,
Christina Lee,
Albert Mitjans Coma,
Romain Moyard,
Mudit Pandey,
Borja Requena,
Matthew Silverman,
Jay Soni,
Antal Száva,
Frederik Wilde,
David Wierichs,
Moritz Willmann.
Published by Jaybsoni almost 2 years ago
Custom measurements can now be facilitated with the addition of the qml.measurements
module. (#3286) (#3343) (#3288) (#3312) (#3287) (#3292) (#3287) (#3326) (#3327) (#3388) (#3439) (#3466)
Within qml.measurements
are new subclasses that allow for the possibility to create custom measurements:
SampleMeasurement
: represents a sample-based measurementStateMeasurement
: represents a state-based measurementMeasurementTransform
: represents a measurement process that requires the application of a batch transformCreating a custom measurement involves making a class that inherits from one of the classes above. An example is given below. Here, the measurement computes the number of samples obtained of a given state:
from pennylane.measurements import SampleMeasurement
class CountState(SampleMeasurement):
def __init__(self, state: str):
self.state = state # string identifying the state, e.g. "0101"
wires = list(range(len(state)))
super().__init__(wires=wires)
def process_samples(self, samples, wire_order, shot_range, bin_size):
counts_mp = qml.counts(wires=self._wires)
counts = counts_mp.process_samples(samples, wire_order, shot_range, bin_size)
return counts.get(self.state, 0)
def __copy__(self):
return CountState(state=self.state)
We can now execute the new measurement in a QNode as follows.
dev = qml.device("default.qubit", wires=1, shots=10000)
@qml.qnode(dev)
def circuit(x):
qml.RX(x, wires=0)
return CountState(state="1")
>>> circuit(1.23)
tensor(3303., requires_grad=True)
Differentiability is also supported for this new measurement process:
>>> x = qml.numpy.array(1.23, requires_grad=True)
>>> qml.grad(circuit)(x)
4715.000000000001
For more information about these new features, see the documentation for qml.measurements
.
ZX diagrams are the medium for which we can envision a quantum circuit as a graph in the ZX-calculus language, showing properties of quantum protocols in a visually compact and logically complete fashion.
QNodes decorated with @qml.transforms.to_zx
will return a PyZX graph that represents the computation in the ZX-calculus language.
dev = qml.device("default.qubit", wires=2)
@qml.transforms.to_zx
@qml.qnode(device=dev)
def circuit(p):
qml.RZ(p[0], wires=1),
qml.RZ(p[1], wires=1),
qml.RX(p[2], wires=0),
qml.PauliZ(wires=0),
qml.RZ(p[3], wires=1),
qml.PauliX(wires=1),
qml.CNOT(wires=[0, 1]),
qml.CNOT(wires=[1, 0]),
qml.SWAP(wires=[0, 1]),
return qml.expval(qml.PauliZ(0) @ qml.PauliZ(1))
>>> params = [5 / 4 * np.pi, 3 / 4 * np.pi, 0.1, 0.3]
>>> circuit(params)
Graph(20 vertices, 23 edges)
Information about PyZX graphs can be found in the PyZX Graphs API.
The symbols and geometry of a compound from the PubChem database can now be accessed via qchem.mol_data()
. (#3289) (#3378)
>>> import pennylane as qml
>>> from pennylane.qchem import mol_data
>>> mol_data("BeH2")
(['Be', 'H', 'H'],
tensor([[ 4.79404621, 0.29290755, 0. ],
[ 3.77945225, -0.29290755, 0. ],
[ 5.80882913, -0.29290755, 0. ]], requires_grad=True))
>>> mol_data(223, "CID")
(['N', 'H', 'H', 'H', 'H'],
tensor([[ 0. , 0. , 0. ],
[ 1.82264085, 0.52836742, 0.40402345],
[ 0.01417295, -1.67429735, -0.98038991],
[-0.98927163, -0.22714508, 1.65369933],
[-0.84773114, 1.373075 , -1.07733286]], requires_grad=True))
Perform quantum chemistry calculations with two new basis sets: 6-311g
and CC-PVDZ
. (#3279)
>>> symbols = ["H", "He"]
>>> geometry = np.array([[1.0, 0.0, 0.0], [0.0, 0.0, 0.0]], requires_grad=False)
>>> charge = 1
>>> basis_names = ["6-311G", "CC-PVDZ"]
>>> for basis_name in basis_names:
... mol = qml.qchem.Molecule(symbols, geometry, charge=charge, basis_name=basis_name)
... print(qml.qchem.hf_energy(mol)())
[-2.84429531]
[-2.84061284]
The controlled CZ gate and controlled Hadamard gate are now available via qml.CCZ
and qml.CH
, respectively. (#3408)
>>> ccz = qml.CCZ(wires=[0, 1, 2])
>>> qml.matrix(ccz)
[[ 1 0 0 0 0 0 0 0]
[ 0 1 0 0 0 0 0 0]
[ 0 0 1 0 0 0 0 0]
[ 0 0 0 1 0 0 0 0]
[ 0 0 0 0 1 0 0 0]
[ 0 0 0 0 0 1 0 0]
[ 0 0 0 0 0 0 1 0]
[ 0 0 0 0 0 0 0 -1]]
>>> ch = qml.CH(wires=[0, 1])
>>> qml.matrix(ch)
[[ 1. 0. 0. 0. ]
[ 0. 1. 0. 0. ]
[ 0. 0. 0.70710678 0.70710678]
[ 0. 0. 0.70710678 -0.70710678]]
Three new parametric operators, qml.CPhaseShift00
, qml.CPhaseShift01
, and qml.CPhaseShift10
, are now available. Each of these operators performs a phase shift akin to qml.ControlledPhaseShift
but on different positions of the state vector. (#2715)
>>> dev = qml.device("default.qubit", wires=2)
>>> @qml.qnode(dev)
>>> def circuit():
... qml.PauliX(wires=1)
... qml.CPhaseShift01(phi=1.23, wires=[0,1])
... return qml.state()
...
>>> circuit()
tensor([0. +0.j , 0.33423773+0.9424888j,
1. +0.j , 0. +0.j ], requires_grad=True)
A new gate operation called qml.FermionicSWAP
has been added. This implements the exchange of spin orbitals representing fermionic-modes while maintaining proper anti-symmetrization. (#3380)
dev = qml.device('default.qubit', wires=2)
@qml.qnode(dev)
def circuit(phi):
qml.BasisState(np.array([0, 1]), wires=[0, 1])
qml.FermionicSWAP(phi, wires=[0, 1])
return qml.state()
>>> circuit(0.1)
tensor([0. +0.j , 0.99750208+0.04991671j,
0.00249792-0.04991671j, 0. +0.j ], requires_grad=True)
Create operators defined from a generator via qml.ops.op_math.Evolution
. (#3375)
qml.ops.op_math.Evolution
defines the exponential of an operator $\hat{O}$ of the form $e^{ix\hat{O}}$, with a single trainable parameter, $x$. Limiting to a single trainable parameter allows the use of qml.gradients.param_shift
to find the gradient with respect to the parameter $x$.
dev = qml.device('default.qubit', wires=2)
@qml.qnode(dev, diff_method=qml.gradients.param_shift)
def circuit(phi):
qml.ops.op_math.Evolution(qml.PauliX(0), -.5 * phi)
return qml.expval(qml.PauliZ(0))
>>> phi = np.array(1.2)
>>> circuit(phi)
tensor(0.36235775, requires_grad=True)
>>> qml.grad(circuit)(phi)
-0.9320390495504149
qml.THadamard
, is now available. (#3340)
The operation accepts a subspace
keyword argument which determines which variant of the qutrit Hadamard to use.
>>> th = qml.THadamard(wires=0, subspace=[0, 1])
>>> qml.matrix(th)
array([[ 0.70710678+0.j, 0.70710678+0.j, 0. +0.j],
[ 0.70710678+0.j, -0.70710678+0.j, 0. +0.j],
[ 0. +0.j, 0. +0.j, 1. +0.j]])
The purity can be calculated in an analogous fashion to, say, the Von Neumann entropy:
qml.math.purity
can be used as an in-line function:
>>> x = [1, 0, 0, 1] / np.sqrt(2)
>>> qml.math.purity(x, [0, 1])
1.0
>>> qml.math.purity(x, [0])
0.5
>>> x = [[1 / 2, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 1 / 2]]
>>> qml.math.purity(x, [0, 1])
0.5
qml.qinfo.transforms.purity
can transform a QNode returning a state to a
function that returns the purity:
dev = qml.device("default.mixed", wires=2)
@qml.qnode(dev)
def circuit(x):
qml.IsingXX(x, wires=[0, 1])
return qml.state()
>>> qml.qinfo.transforms.purity(circuit, wires=[0])(np.pi / 2)
0.5
>>> qml.qinfo.transforms.purity(circuit, wires=[0, 1])(np.pi / 2)
1.0
As with the other methods in qml.qinfo
, the purity is fully differentiable:
>>> param = np.array(np.pi / 4, requires_grad=True)
>>> qml.grad(qml.qinfo.transforms.purity(circuit, wires=[0]))(param)
-0.5
qml.gradients.spsa_grad
, that is based on the idea of SPSA is now available. (#3366)
This new transform allows users to compute a single estimate of a quantum gradient using simultaneous perturbation of parameters and a stochastic approximation. A QNode that takes, say, an argument x
, the approximate gradient can be computed as follows.
>>> dev = qml.device("default.qubit", wires=2)
>>> x = np.array(0.4, requires_grad=True)
>>> @qml.qnode(dev)
... def circuit(x):
... qml.RX(x, 0)
... qml.RX(x, 1)
... return qml.expval(qml.PauliZ(0))
>>> grad_fn = qml.gradients.spsa_grad(circuit, h=0.1, num_directions=1)
>>> grad_fn(x)
array(-0.38876964)
The argument num_directions
determines how many directions of simultaneous perturbation are used, which is proportional to the number of circuit evaluations. See the SPSA gradient transform documentation for details. Note that the full SPSA optimizer is already available as qml.SPSAOptimizer
.
Multiple mid-circuit measurements can now be combined arithmetically to create new conditionals. (#3159)
dev = qml.device("default.qubit", wires=3)
@qml.qnode(dev)
def circuit():
qml.Hadamard(wires=0)
qml.Hadamard(wires=1)
m0 = qml.measure(wires=0)
m1 = qml.measure(wires=1)
combined = 2 * m1 + m0
qml.cond(combined == 2, qml.RX)(1.3, wires=2)
return qml.probs(wires=2)
>>> circuit()
[0.90843735 0.09156265]
A new method called pauli_decompose()
has been added to the qml.pauli
module, which takes a hermitian matrix, decomposes it in the Pauli basis, and returns it either as a qml.Hamiltonian
or qml.PauliSentence
instance. (#3384)
Operation
or Hamiltonian
instances can now be generated from a qml.PauliSentence
or qml.PauliWord
via the new operation()
and hamiltonian()
methods. (#3391)
>>> pw = qml.pauli.PauliWord({0: 'X', 1: 'Y'})
>>> print(pw.operation())
PauliX(wires=[0]) @ PauliY(wires=[1])
>>> print(pw.hamiltonian())
(1) [X0 Y1]
>>> ps = qml.pauli.PauliSentence({pw: -1.23})
>>> print(ps.operation())
-1.23*(PauliX(wires=[0]) @ PauliY(wires=[1]))
>>> print(ps.hamiltonian())
(-1.23) [X0 Y1]
A sum_expand
function has been added for tapes, which splits a tape measuring a Sum
expectation into mutliple tapes of summand expectations, and provides a function to recombine the results. (#3230)
The autograd and Tensorflow interfaces now support devices with shot vectors when qml.enable_return()
has been called. (#3374) (#3400)
Here is an example using Tensorflow:
import tensorflow as tf
qml.enable_return()
dev = qml.device("default.qubit", wires=2, shots=[1000, 2000, 3000])
@qml.qnode(dev, diff_method="parameter-shift", interface="tf")
def circuit(a):
qml.RY(a, wires=0)
qml.RX(0.2, wires=0)
qml.CNOT(wires=[0, 1])
return qml.expval(qml.PauliZ(0)), qml.probs([0, 1])
>>> a = tf.Variable(0.4)
>>> with tf.GradientTape() as tape:
... res = circuit(a)
... res = tf.stack([tf.experimental.numpy.hstack(r) for r in res])
...
>>> res
<tf.Tensor: shape=(3, 5), dtype=float64, numpy=
array([[0.902, 0.951, 0. , 0. , 0.049],
[0.898, 0.949, 0. , 0. , 0.051],
[0.892, 0.946, 0. , 0. , 0.054]])>
>>> tape.jacobian(res, a)
<tf.Tensor: shape=(3, 5), dtype=float64, numpy=
array([[-0.345 , -0.1725 , 0. , 0. , 0.1725 ],
[-0.383 , -0.1915 , 0. , 0. , 0.1915 ],
[-0.38466667, -0.19233333, 0. , 0. , 0.19233333]])>
The PyTorch interface is now fully supported when qml.enable_return()
has been called, allowing the calculation of the Jacobian and the Hessian using custom differentiation methods (e.g., parameter-shift, finite difference, or adjoint). (#3416)
import torch
qml.enable_return()
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev, diff_method="parameter-shift", interface="torch")
def circuit(a, b):
qml.RY(a, wires=0)
qml.RX(b, wires=1)
qml.CNOT(wires=[0, 1])
return qml.expval(qml.PauliZ(0)), qml.probs([0, 1])
>>> a = torch.tensor(0.1, requires_grad=True)
>>> b = torch.tensor(0.2, requires_grad=True)
>>> torch.autograd.functional.jacobian(circuit, (a, b))
((tensor(-0.0998), tensor(0.)), (tensor([-0.0494, -0.0005, 0.0005, 0.0494]), tensor([-0.0991, 0.0991, 0.0002, -0.0002])))
The JAX-JIT interface now supports first-order gradient computation when qml.enable_return()
has been called. (#3235) (#3445)
import jax
from jax import numpy as jnp
jax.config.update("jax_enable_x64", True)
qml.enable_return()
dev = qml.device("lightning.qubit", wires=2)
@jax.jit
@qml.qnode(dev, interface="jax-jit", diff_method="parameter-shift")
def circuit(a, b):
qml.RY(a, wires=0)
qml.RX(b, wires=0)
return qml.expval(qml.PauliZ(0)), qml.expval(qml.PauliZ(1))
a, b = jnp.array(1.0), jnp.array(2.0)
>>> jax.jacobian(circuit, argnums=[0, 1])(a, b)
((Array(0.35017549, dtype=float64, weak_type=True),
Array(-0.4912955, dtype=float64, weak_type=True)),
(Array(5.55111512e-17, dtype=float64, weak_type=True),
Array(0., dtype=float64, weak_type=True)))
qml.pauli.is_pauli_word
now supports instances of qml.Hamiltonian
. (#3389)
When qml.probs
, qml.counts
, and qml.sample
are called with no arguments, they measure all wires. Calling any of the aforementioned measurements with an empty wire list (e.g., qml.sample(wires=[])
) will raise an error. (#3299)
Made qml.gradients.finite_diff
more convenient to use with custom data type observables/devices by reducing the number of magic methods that need to be defined in the custom data type to support finite_diff
. (#3426)
The qml.ISWAP
gate is now natively supported on default.mixed
, improving on its efficiency. (#3284)
Added more input validation to qml.transforms.hamiltonian_expand
such that Hamiltonian objects with no terms raise an error. (#3339)
Continuous integration checks are now performed for Python 3.11 and Torch v1.13. Python 3.7 is dropped. (#3276)
qml.Tracker
now also logs results in tracker.history
when tracking the execution of a circuit. (#3306)
The execution time of Wires.all_wires
has been improved by avoiding data type changes and making use of itertools.chain
. (#3302)
Printing an instance of qml.qchem.Molecule
is now more concise and informational. (#3364)
The error message for qml.transforms.insert
when it fails to diagonalize non-qubit-wise-commuting observables is now more detailed. (#3381)
Extended the qml.equal
function to qml.Hamiltonian
and Tensor
objects. (#3390)
QuantumTape._process_queue
has been moved to qml.queuing.process_queue
to disentangle its functionality from the QuantumTape
class. (#3401)
QPE can now accept a target operator instead of a matrix and target wires pair. (#3373)
The qml.ops.op_math.Controlled.map_wires
method now uses base.map_wires
internally instead of the private _wires
property setter. (#3405)
A new function called qml.tape.make_qscript
has been created for converting a quantum function into a quantum script. This replaces qml.transforms.make_tape
. (#3429)
Add a _pauli_rep
attribute to operators to integrate the new Pauli arithmetic classes with native PennyLane objects. (#3443)
Extended the functionality of qml.matrix
to qutrits. (#3508)
The qcut.py
file in pennylane/transforms/
has been reorganized into multiple files that are now in pennylane/transforms/qcut/
. (#3413)
A warning now appears when creating a Tensor
object with overlapping wires, informing that this can lead to undefined behaviour. (#3459)
Extended the qml.equal
function to qml.ops.op_math.Controlled
and qml.ops.op_math.ControlledOp
objects. (#3463)
Nearly every instance of with QuantumTape()
has been replaced with QuantumScript
construction. (#3454)
Added validate_subspace
static method to qml.Operator
to check the validity of the subspace of certain
qutrit operations. (#3340)
qml.equal
now supports operators created via qml.s_prod
, qml.pow
, qml.exp
, and qml.adjoint
. (#3471)
Devices can now disregard observable grouping indices in Hamiltonians through the optional use_grouping
attribute. (#3456)
Add the optional argument lazy=True
to functions qml.s_prod
, qml.prod
and qml.op_sum
to allow simplification. (#3483)
Updated the qml.transforms.zyz_decomposition
function such that it now supports broadcast operators. This means that single-qubit qml.QubitUnitary
operators, instantiated from a batch of unitaries, can now be decomposed. (#3477)
The performance of executing circuits under the jax.vmap
transformation has been improved by being able to leverage the batch-execution capabilities of some devices. (#3452)
The tolerance for converting openfermion Hamiltonian complex coefficients to real ones has been modified to prevent conversion errors. (#3367)
OperationRecorder
now inherits from AnnotatedQueue
and QuantumScript
instead of QuantumTape
. (#3496)
Updated qml.transforms.split_non_commuting
to support the new return types. (#3414)
Updated qml.transforms.mitigate_with_zne
to support the new return types. (#3415)
Updated qml.transforms.metric_tensor
, qml.transforms.adjoint_metric_tensor
, qml.qinfo.classical_fisher
, and qml.qinfo.quantum_fisher
to support the new return types. (#3449)
Updated qml.transforms.batch_params
and qml.transforms.batch_input
to support the new return types. (#3431)
Updated qml.transforms.cut_circuit
and qml.transforms.cut_circuit_mc
to support the new return types. (#3346)
Limit NumPy version to <1.24
. (#3346)
Python 3.7 support is no longer maintained. PennyLane will be maintained for versions 3.8 and up. (#3276)
The log_base
attribute has been moved from MeasurementProcess
to the new VnEntropyMP
and MutualInfoMP
classes, which inherit from MeasurementProcess
. (#3326)
qml.utils.decompose_hamiltonian()
has been removed. Please use qml.pauli.pauli_decompose()
instead. (#3384)
The return_type
attribute of MeasurementProcess
has been removed where possible. Use isinstance
checks instead. (#3399)
Instead of having an OrderedDict
attribute called _queue
, AnnotatedQueue
now inherits from OrderedDict
and encapsulates the queue. Consequentially, this also applies to the QuantumTape
class which inherits from AnnotatedQueue
. (#3401)
The ShadowMeasurementProcess
class has been renamed to ClassicalShadowMP
. (#3388)
The qml.Operation.get_parameter_shift
method has been removed. The gradients
module should be used for general parameter-shift rules instead. (#3419)
The signature of the QubitDevice.statistics
method has been changed from
def statistics(self, observables, shot_range=None, bin_size=None, circuit=None):
to
def statistics(self, circuit: QuantumTape, shot_range=None, bin_size=None):
MeasurementProcess
class is now an abstract class and return_type
is now a property of the class. (#3434)
Deprecations cycles are tracked at doc/developement/deprecations.rst.
The following methods are deprecated: (#3281)
qml.tape.get_active_tape
: Use qml.QueuingManager.active_context()
insteadqml.transforms.qcut.remap_tape_wires
: Use qml.map_wires
insteadqml.tape.QuantumTape.inv()
: Use qml.tape.QuantumTape.adjoint()
insteadqml.tape.stop_recording()
: Use qml.QueuingManager.stop_recording()
insteadqml.tape.QuantumTape.stop_recording()
: Use qml.QueuingManager.stop_recording()
insteadqml.QueuingContext
is now qml.QueuingManager
QueuingManager.safe_update_info
and AnnotatedQueue.safe_update_info
: Use update_info
instead.qml.transforms.measurement_grouping
has been deprecated. Use qml.transforms.hamiltonian_expand
instead. (#3417)
The observables
argument in QubitDevice.statistics
is deprecated. Please use circuit
instead. (#3433)
The seed_recipes
argument in qml.classical_shadow
and qml.shadow_expval
is deprecated. A new argument seed
has been added, which defaults to None and can contain an integer with the wanted seed. (#3388)
qml.transforms.make_tape
has been deprecated. Please use qml.tape.make_qscript
instead. (#3478)
Added documentation on parameter broadcasting regarding both its usage and technical aspects. (#3356)
The quickstart guide on circuits as well as the the documentation of QNodes and Operators now contain introductions and details on parameter broadcasting. The QNode documentation mostly contains usage details, the Operator documentation is concerned with implementation details and a guide to support broadcasting in custom operators.
The return type statements of gradient and Hessian transforms and a series of other functions that are a batch_transform
have been corrected. (#3476)
Developer documentation for the queuing module has been added. (#3268)
More mentions of diagonalizing gates for all relevant operations have been corrected. (#3409)
The docstrings for compute_eigvals
used to say that the diagonalizing gates implemented $U$, the unitary such that $O = U \Sigma U^{\dagger}$, where $O$ is the original observable and $\Sigma$ a diagonal matrix. However, the diagonalizing gates actually implement $U^{\dagger}$, since $\langle \psi | O | \psi \rangle = \langle \psi | U \Sigma U^{\dagger} | \psi \rangle$, making $U^{\dagger} | \psi \rangle$ the actual state being measured in the $Z$-basis.
A warning about using dill
to pickle and unpickle datasets has been added. (#3505)
Fixed a bug that prevented qml.gradients.param_shift
from being used for broadcasted tapes. (#3528)
Fixed a bug where qml.transforms.hamiltonian_expand
didn't preserve the type of the input results in its output. (#3339)
Fixed a bug that made qml.gradients.param_shift
raise an error when used with unshifted terms only in a custom recipe, and when using any unshifted terms at all under the new return type system. (#3177)
The original tape _obs_sharing_wires
attribute is updated during its expansion. (#3293)
An issue with drain=False
in the adaptive optimizer has been fixed. Before the fix, the operator pool needed to be reconstructed inside the optimization pool when drain=False
. With this fix, this reconstruction is no longer needed. (#3361)
If the device originally has no shots but finite shots are dynamically specified, Hamiltonian expansion now occurs. (#3369)
qml.matrix(op)
now fails if the operator truly has no matrix (e.g., qml.Barrier
) to match op.matrix()
. (#3386)
The pad_with
argument in the qml.AmplitudeEmbedding
template is now compatible with all interfaces. (#3392)
Operator.pow
now queues its constituents by default. (#3373)
Fixed a bug where a QNode returning qml.sample
would produce incorrect results when run on a device defined with a shot vector. (#3422)
The qml.data
module now works as expected on Windows. (#3504)
This release contains contributions from (in alphabetical order):
Guillermo Alonso, Juan Miguel Arrazola, Utkarsh Azad, Samuel Banning, Thomas Bromley, Astral Cai, Albert Mitjans Coma, Ahmed Darwish, Isaac De Vlugt, Olivia Di Matteo, Amintor Dusko, Pieter Eendebak, Lillian M. A. Frederiksen, Diego Guala, Katharine Hyatt, Josh Izaac, Soran Jahangiri, Edward Jiang, Korbinian Kottmann, Christina Lee, Romain Moyard, Lee James O'Riordan, Mudit Pandey, Kevin Shen, Matthew Silverman, Jay Soni, Antal Száva, David Wierichs, Moritz Willmann, and Filippo Vicentini.
Published by Jaybsoni almost 2 years ago
The qml.data
module is now available, allowing users to download, load, and create quantum datasets. (#3156)
Datasets are hosted on Xanadu Cloud and can be downloaded by using qml.data.load()
:
>>> H2_datasets = qml.data.load(
... data_name="qchem", molname="H2", basis="STO-3G", bondlength=1.1
... )
>>> H2data = H2_datasets[0]
>>> H2data
<Dataset = description: qchem/H2/STO-3G/1.1, attributes: ['molecule', 'hamiltonian', ...]>
Datasets available to be downloaded can be listed with qml.data.list_datasets()
.
To download or load only specific properties of a dataset, we can specify the desired properties in qml.data.load
with the attributes
keyword argument:
>>> H2_hamiltonian = qml.data.load(
... data_name="qchem", molname="H2", basis="STO-3G", bondlength=1.1,
... attributes=["molecule", "hamiltonian"]
... )[0]
>>> H2_hamiltonian.hamiltonian
<Hamiltonian: terms=15, wires=[0, 1, 2, 3]>
The available attributes can be found using qml.data.list_attributes()
:
To select data interactively, we can use qml.data.load_interactive()
:
>>> qml.data.load_interactive()
Please select a data name:
1) qspin
2) qchem
Choice [1-2]: 1
Please select a sysname:
...
Please select a periodicity:
...
Please select a lattice:
...
Please select a layout:
...
Please select attributes:
...
Force download files? (Default is no) [y/N]: N
Folder to download to? (Default is pwd, will download to /datasets subdirectory):
Please confirm your choices:
dataset: qspin/Ising/open/rectangular/4x4
attributes: ['parameters', 'ground_states']
force: False
dest folder: datasets
Would you like to continue? (Default is yes) [Y/n]:
<Dataset = description: qspin/Ising/open/rectangular/4x4, attributes: ['parameters', 'ground_states']>
Once a dataset is loaded, its properties can be accessed as follows:
>>> dev = qml.device("default.qubit",wires=4)
>>> @qml.qnode(dev)
... def circuit():
... qml.BasisState(H2data.hf_state, wires = [0, 1, 2, 3])
... for op in H2data.vqe_gates:
... qml.apply(op)
... return qml.expval(H2data.hamiltonian)
>>> print(circuit())
-1.0791430411076344
It's also possible to create custom datasets with qml.data.Dataset
:
>>> example_hamiltonian = qml.Hamiltonian(coeffs=[1,0.5], observables=[qml.PauliZ(wires=0),qml.PauliX(wires=1)])
>>> example_energies, _ = np.linalg.eigh(qml.matrix(example_hamiltonian))
>>> example_dataset = qml.data.Dataset(
... data_name = 'Example', hamiltonian=example_hamiltonian, energies=example_energies
... )
>>> example_dataset.data_name
'Example'
>>> example_dataset.hamiltonian
(0.5) [X1]
+ (1) [Z0]
>>> example_dataset.energies
array([-1.5, -0.5, 0.5, 1.5])
Custom datasets can be saved and read with the qml.data.Dataset.write()
and qml.data.Dataset.read()
methods, respectively.
>>> example_dataset.write('./path/to/dataset.dat')
>>> read_dataset = qml.data.Dataset()
>>> read_dataset.read('./path/to/dataset.dat')
>>> read_dataset.data_name
'Example'
>>> read_dataset.hamiltonian
(0.5) [X1]
+ (1) [Z0]
>>> read_dataset.energies
array([-1.5, -0.5, 0.5, 1.5])
We will continue to work on adding more datasets and features for qml.data
in future releases.
Optimizing quantum circuits can now be done adaptively with qml.AdaptiveOptimizer
. (#3192)
The qml.AdaptiveOptimizer
takes an initial circuit and a collection of operators as input and adds a selected gate to the circuit at each optimization step. The process of growing the circuit can be repeated until the circuit gradients converge to zero within a given threshold. The adaptive optimizer can be used to implement algorithms such as ADAPT-VQE as shown in the following example.
Firstly, we define some preliminary variables needed for VQE:
symbols = ["H", "H", "H"]
geometry = np.array([[0.01076341, 0.04449877, 0.0],
[0.98729513, 1.63059094, 0.0],
[1.87262415, -0.00815842, 0.0]], requires_grad=False)
H, qubits = qml.qchem.molecular_hamiltonian(symbols, geometry, charge = 1)
The collection of gates to grow the circuit is built to contain all single and double excitations:
n_electrons = 2
singles, doubles = qml.qchem.excitations(n_electrons, qubits)
singles_excitations = [qml.SingleExcitation(0.0, x) for x in singles]
doubles_excitations = [qml.DoubleExcitation(0.0, x) for x in doubles]
operator_pool = doubles_excitations + singles_excitations
Next, an initial circuit that prepares a Hartree-Fock state and returns the expectation value of the Hamiltonian is defined:
hf_state = qml.qchem.hf_state(n_electrons, qubits)
dev = qml.device("default.qubit", wires=qubits)
@qml.qnode(dev)
def circuit():
qml.BasisState(hf_state, wires=range(qubits))
return qml.expval(H)
Finally, the optimizer is instantiated and then the circuit is created and optimized adaptively:
opt = qml.optimize.AdaptiveOptimizer()
for i in range(len(operator_pool)):
circuit, energy, gradient = opt.step_and_cost(circuit, operator_pool, drain_pool=True)
print('Energy:', energy)
print(qml.draw(circuit)())
print('Largest Gradient:', gradient)
print()
if gradient < 1e-3:
break
Energy: -1.246549938420637
0: ─╭BasisState(M0)─╭G²(0.20)─┤ ╭<𝓗>
1: ─├BasisState(M0)─├G²(0.20)─┤ ├<𝓗>
2: ─├BasisState(M0)─│─────────┤ ├<𝓗>
3: ─├BasisState(M0)─│─────────┤ ├<𝓗>
4: ─├BasisState(M0)─├G²(0.20)─┤ ├<𝓗>
5: ─╰BasisState(M0)─╰G²(0.20)─┤ ╰<𝓗>
Largest Gradient: 0.14399872776755085
Energy: -1.2613740231529604
0: ─╭BasisState(M0)─╭G²(0.20)─╭G²(0.19)─┤ ╭<𝓗>
1: ─├BasisState(M0)─├G²(0.20)─├G²(0.19)─┤ ├<𝓗>
2: ─├BasisState(M0)─│─────────├G²(0.19)─┤ ├<𝓗>
3: ─├BasisState(M0)─│─────────╰G²(0.19)─┤ ├<𝓗>
4: ─├BasisState(M0)─├G²(0.20)───────────┤ ├<𝓗>
5: ─╰BasisState(M0)─╰G²(0.20)───────────┤ ╰<𝓗>
Largest Gradient: 0.1349349562423238
Energy: -1.2743971719780331
0: ─╭BasisState(M0)─╭G²(0.20)─╭G²(0.19)──────────┤ ╭<𝓗>
1: ─├BasisState(M0)─├G²(0.20)─├G²(0.19)─╭G(0.00)─┤ ├<𝓗>
2: ─├BasisState(M0)─│─────────├G²(0.19)─│────────┤ ├<𝓗>
3: ─├BasisState(M0)─│─────────╰G²(0.19)─╰G(0.00)─┤ ├<𝓗>
4: ─├BasisState(M0)─├G²(0.20)────────────────────┤ ├<𝓗>
5: ─╰BasisState(M0)─╰G²(0.20)────────────────────┤ ╰<𝓗>
Largest Gradient: 0.00040841755397108586
For a detailed breakdown of its implementation, check out the Adaptive circuits for quantum chemistry demo.
QNodes now accept an auto
interface argument which automatically detects the machine learning library to use. (#3132)
from pennylane import numpy as np
import torch
import tensorflow as tf
from jax import numpy as jnp
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev, interface="auto")
def circuit(weight):
qml.RX(weight[0], wires=0)
qml.RY(weight[1], wires=1)
return qml.expval(qml.PauliZ(0))
interface_tensors = [[0, 1], np.array([0, 1]), torch.Tensor([0, 1]), tf.Variable([0, 1], dtype=float), jnp.array([0, 1])]
for tensor in interface_tensors:
res = circuit(weight=tensor)
print(f"Result value: {res:.2f}; Result type: {type(res)}")
Result value: 1.00; Result type: <class 'pennylane.numpy.tensor.tensor'>
Result value: 1.00; Result type: <class 'pennylane.numpy.tensor.tensor'>
Result value: 1.00; Result type: <class 'torch.Tensor'>
Result value: 1.00; Result type: <class 'tensorflow.python.framework.ops.EagerTensor'>
Result value: 1.00; Result type: <class 'jaxlib.xla_extension.DeviceArray'>
JAX-JIT support for computing the gradient of QNodes that return a single vector of probabilities or multiple expectation values is now available. (#3244) (#3261)
import jax
from jax import numpy as jnp
from jax.config import config
config.update("jax_enable_x64", True)
dev = qml.device("lightning.qubit", wires=2)
@jax.jit
@qml.qnode(dev, diff_method="parameter-shift", interface="jax")
def circuit(x, y):
qml.RY(x, wires=0)
qml.RY(y, wires=1)
qml.CNOT(wires=[0, 1])
return qml.expval(qml.PauliZ(0)), qml.expval(qml.PauliZ(1))
x = jnp.array(1.0)
y = jnp.array(2.0)
>>> jax.jacobian(circuit, argnums=[0, 1])(x, y)
(DeviceArray([-0.84147098, 0.35017549], dtype=float64, weak_type=True),
DeviceArray([ 4.47445479e-18, -4.91295496e-01], dtype=float64, weak_type=True))
Note that this change depends on jax.pure_callback
, which requires jax>=0.3.17
.
We've reorganized and grouped everything in PennyLane responsible for manipulating Pauli operators into a pauli
module. The grouping
module has been deprecated as a result, and logic was moved from pennylane/grouping
to pennylane/pauli/grouping
. (#3179)
qml.pauli.PauliWord
and qml.pauli.PauliSentence
can be used to represent tensor products and linear combinations of Pauli operators, respectively. These provide a more performant method to compute sums and products of Pauli operators. (#3195)
qml.pauli.PauliWord
represents tensor products of Pauli operators. We can efficiently multiply and extract the matrix of these operators using this representation.
>>> pw1 = qml.pauli.PauliWord({0:"X", 1:"Z"})
>>> pw2 = qml.pauli.PauliWord({0:"Y", 1:"Z"})
>>> pw1, pw2
(X(0) @ Z(1), Y(0) @ Z(1))
>>> pw1 * pw2
(Z(0), 1j)
>>> pw1.to_mat(wire_order=[0,1])
array([[ 0, 0, 1, 0],
[ 0, 0, 0, -1],
[ 1, 0, 0, 0],
[ 0, -1, 0, 0]])
qml.pauli.PauliSentence
represents linear combinations of Pauli words. We can efficiently add, multiply and extract the matrix of these operators in this representation.
>>> ps1 = qml.pauli.PauliSentence({pw1: 1.2, pw2: 0.5j})
>>> ps2 = qml.pauli.PauliSentence({pw1: -1.2})
>>> ps1
1.2 * X(0) @ Z(1)
+ 0.5j * Y(0) @ Z(1)
>>> ps1 + ps2
0.0 * X(0) @ Z(1)
+ 0.5j * Y(0) @ Z(1)
>>> ps1 * ps2
-1.44 * I
+ (-0.6+0j) * Z(0)
>>> (ps1 + ps2).to_mat(wire_order=[0,1])
array([[ 0. +0.j, 0. +0.j, 0.5+0.j, 0. +0.j],
[ 0. +0.j, 0. +0.j, 0. +0.j, -0.5+0.j],
[-0.5+0.j, 0. +0.j, 0. +0.j, 0. +0.j],
[ 0. +0.j, 0.5+0.j, 0. +0.j, 0. +0.j]])
qml.enable_return()
now supports QNodes returning multiple measurements, including shots vectors, and gradient output types. (#2886) (#3052) (#3041) (#3090) (#3069) (#3137) (#3127) (#3099) (#3098) (#3095) (#3091) (#3176) (#3170) (#3194) (#3267) (#3234) (#3232) (#3223) (#3222) (#3315)
In v0.25, we introduced qml.enable_return()
, which separates measurements into their own tensors. The motivation of this change is the deprecation of ragged ndarray
creation in NumPy.
With this release, we're continuing to elevate this feature by adding support for:
Execution (qml.execute
)
Jacobian vector product (JVP) computation
Gradient transforms (qml.gradients.param_shift
, qml.gradients.finite_diff
, qml.gradients.hessian_transform
, qml.gradients.param_shift_hessian
).
Interfaces (Autograd, TensorFlow, and JAX, although without JIT)
With this added support, the JAX interface can handle multiple shots (shots vectors), measurements, and gradient output types with qml.enable_return()
:
import jax
qml.enable_return()
dev = qml.device("default.qubit", wires=2, shots=(1, 10000))
params = jax.numpy.array([0.1, 0.2])
@qml.qnode(dev, interface="jax", diff_method="parameter-shift", max_diff=2)
def circuit(x):
qml.RX(x[0], wires=[0])
qml.RY(x[1], wires=[1])
qml.CNOT(wires=[0, 1])
return qml.var(qml.PauliZ(0) @ qml.PauliX(1)), qml.probs(wires=[0])
>>> jax.hessian(circuit)(params)
((DeviceArray([[ 0., 0.],
[ 2., -3.]], dtype=float32),
DeviceArray([[[-0.5, 0. ],
[ 0. , 0. ]],
[[ 0.5, 0. ],
[ 0. , 0. ]]], dtype=float32)),
(DeviceArray([[ 0.07677898, 0.0563341 ],
[ 0.07238522, -1.830669 ]], dtype=float32),
DeviceArray([[[-4.9707499e-01, 2.9999996e-04],
[-6.2500127e-04, 1.2500001e-04]],
[[ 4.9707499e-01, -2.9999996e-04],
[ 6.2500127e-04, -1.2500001e-04]]], dtype=float32)))
For more details, please refer to the documentation.
Grouped coefficients, observables, and basis rotation transformation matrices needed to construct a qubit Hamiltonian in the rotated basis of molecular orbitals are now calculable via qml.qchem.basis_rotation()
. (#3011)
>>> symbols = ['H', 'H']
>>> geometry = np.array([[0.0, 0.0, 0.0], [1.398397361, 0.0, 0.0]], requires_grad = False)
>>> mol = qml.qchem.Molecule(symbols, geometry)
>>> core, one, two = qml.qchem.electron_integrals(mol)()
>>> coeffs, ops, unitaries = qml.qchem.basis_rotation(one, two, tol_factor=1.0e-5)
>>> unitaries
[tensor([[-1.00000000e+00, -5.46483514e-13],
[ 5.46483514e-13, -1.00000000e+00]], requires_grad=True),
tensor([[-1.00000000e+00, 3.17585063e-14],
[-3.17585063e-14, -1.00000000e+00]], requires_grad=True),
tensor([[-0.70710678, -0.70710678],
[-0.70710678, 0.70710678]], requires_grad=True),
tensor([[ 2.58789009e-11, 1.00000000e+00],
[-1.00000000e+00, 2.58789009e-11]], requires_grad=True)]
Any gate operation can now be tapered according to :math:\mathbb{Z}_2
symmetries of the Hamiltonian via qml.qchem.taper_operation
. (#3002) (#3121)
>>> symbols = ['He', 'H']
>>> geometry = np.array([[0.0, 0.0, 0.0], [0.0, 0.0, 1.4589]])
>>> mol = qml.qchem.Molecule(symbols, geometry, charge=1)
>>> H, n_qubits = qml.qchem.molecular_hamiltonian(symbols, geometry)
>>> generators = qml.qchem.symmetry_generators(H)
>>> paulixops = qml.qchem.paulix_ops(generators, n_qubits)
>>> paulix_sector = qml.qchem.optimal_sector(H, generators, mol.n_electrons)
>>> tap_op = qml.qchem.taper_operation(qml.SingleExcitation, generators, paulixops,
... paulix_sector, wire_order=H.wires, op_wires=[0, 2])
>>> tap_op(3.14159)
[Exp(1.5707949999999993j PauliY)]
Moreover, the obtained tapered operation can be used directly within a QNode.
>>> dev = qml.device('default.qubit', wires=[0, 1])
>>> @qml.qnode(dev)
... def circuit(params):
... tap_op(params[0])
... return qml.expval(qml.PauliZ(0) @ qml.PauliZ(1))
>>> drawer = qml.draw(circuit, show_all_wires=True)
>>> print(drawer(params=[3.14159]))
0: ──Exp(0.00+1.57j Y)─┤ ╭<Z@Z>
1: ────────────────────┤ ╰<Z@Z>
Functionality has been added to estimate the number of measurements required to compute an expectation value with a target error and estimate the error in computing an expectation value with a given number of measurements. (#3000)
Wires of operators or entire QNodes can now be mapped to other wires via qml.map_wires()
. (#3143) (#3145)
The qml.map_wires()
function requires a dictionary representing a wire map. Use it with
arbitrary operators:
>>> op = qml.RX(0.54, wires=0) + qml.PauliX(1) + (qml.PauliZ(2) @ qml.RY(1.23, wires=3))
>>> op
(RX(0.54, wires=[0]) + PauliX(wires=[1])) + (PauliZ(wires=[2]) @ RY(1.23, wires=[3]))
>>> wire_map = {0: 10, 1: 11, 2: 12, 3: 13}
>>> qml.map_wires(op, wire_map)
(RX(0.54, wires=[10]) + PauliX(wires=[11])) + (PauliZ(wires=[12]) @ RY(1.23, wires=[13]))
A map_wires
method has also been added to operators, which returns a copy
of the operator with its wires changed according to the given wire map.
entire QNodes:
dev = qml.device("default.qubit", wires=["A", "B", "C", "D"])
wire_map = {0: "A", 1: "B", 2: "C", 3: "D"}
@qml.qnode(dev)
def circuit():
qml.RX(0.54, wires=0)
qml.PauliX(1)
qml.PauliZ(2)
qml.RY(1.23, wires=3)
return qml.probs(wires=0)
>>> mapped_circuit = qml.map_wires(circuit, wire_map)
>>> mapped_circuit()
tensor([0.92885434, 0.07114566], requires_grad=True)
>>> print(qml.draw(mapped_circuit)())
A: ──RX(0.54)─┤ Probs
B: ──X────────┤
C: ──Z────────┤
D: ──RY(1.23)─┤
The qml.IntegerComparator
arithmetic operation is now available. (#3113)
Given a basis state :math:\vert n \rangle
, where :math:n
is a positive integer, and a fixed positive integer :math:L
, qml.IntegerComparator
flips a target qubit if :math:n \geq L
. Alternatively, the flipping condition can be :math:n < L
as demonstrated below:
dev = qml.device("default.qubit", wires=2)
@qml.qnode(dev)
def circuit():
qml.BasisState(np.array([0, 1]), wires=range(2))
qml.broadcast(qml.Hadamard, wires=range(2), pattern='single')
qml.IntegerComparator(2, geq=False, wires=[0, 1])
return qml.state()
>>> circuit()
[-0.5+0.j 0.5+0.j -0.5+0.j 0.5+0.j]
The qml.GellMann
qutrit observable, the ternary generalization of the Pauli observables, is now available. (#3035)
When using qml.GellMann
, the index
keyword argument determines which of the 8 Gell-Mann matrices is used.
dev = qml.device("default.qutrit", wires=2)
@qml.qnode(dev)
def circuit():
qml.TClock(wires=0)
qml.TShift(wires=1)
qml.TAdd(wires=[0, 1])
return qml.expval(qml.GellMann(wires=0, index=8) + qml.GellMann(wires=1, index=3))
>>> circuit()
-0.42264973081037416
Controlled qutrit operations can now be performed with qml.ControlledQutritUnitary
. (#2844)
The control wires and values that define the operation are defined analogously to the qubit operation.
dev = qml.device("default.qutrit", wires=3)
@qml.qnode(dev)
def circuit(U):
qml.TShift(wires=0)
qml.TAdd(wires=[0, 1])
qml.ControlledQutritUnitary(U, control_wires=[0, 1], control_values='12', wires=2)
return qml.state()
>>> U = np.array([[1, 1, 0], [1, -1, 0], [0, 0, np.sqrt(2)]]) / np.sqrt(2)
>>> circuit(U)
tensor([0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j,
0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 1.+0.j, 0.+0.j, 0.+0.j, 0.+0.j,
0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j, 0.+0.j,
0.+0.j, 0.+0.j, 0.+0.j], requires_grad=True)
PennyLane now supports Python 3.11! (#3297)
qml.sample
and qml.counts
work more efficiently and track if computational basis samples are being generated when they are called without specifying an observable. (#3207)
The parameters of a basis set containing a different number of Gaussian functions are now easier to differentiate. (#3213)
Printing a qml.MultiControlledX
operator now shows the control_values
keyword argument. (#3113)
qml.simplify
and transforms like qml.matrix
, batch_transform
, hamiltonian_expand
, and split_non_commuting
now work with QuantumScript
as well as QuantumTape
. (#3209)
A redundant flipping of the initial state in the UCCSD and kUpCCGSD templates has been removed. (#3148)
qml.adjoint
now supports batching if the base operation supports batching. (#3168)
qml.OrbitalRotation
is now decomposed into two qml.SingleExcitation
operations for faster execution and more efficient parameter-shift gradient calculations on devices that natively support qml.SingleExcitation
. (#3171)
The Exp
class decomposes into a PauliRot
class if the coefficient is imaginary and the base operator is a Pauli Word. (#3249)
Added the operator attributes has_decomposition
and has_adjoint
that indicate whether a corresponding decomposition
or adjoint
method is available. (#2986)
Structural improvements are made to QueuingManager
, formerly QueuingContext
, and AnnotatedQueue
. (#2794) (#3061) (#3085)
QueuingContext
is renamed to QueuingManager
.QueuingManager
should now be the global communication point for putting queuable objects into the active queue.QueuingManager
is no longer an abstract base class.AnnotatedQueue
and its children no longer inherit from QueuingManager
.QueuingManager
is no longer a context manager.QueuingManager.add_active_queue
and QueuingContext.remove_active_queue
class methods instead of directly manipulating the _active_contexts
property.AnnotatedQueue
and its children no longer provide global information about actively recording queues. This information is now only available through QueuingManager
.AnnotatedQueue
and its children no longer have the private _append
, _remove
, _update_info
, _safe_update_info
, and _get_info
methods. The public analogues should be used instead.QueuingManager.safe_update_info
and AnnotatedQueue.safe_update_info
are deprecated. Their functionality is moved to update_info
.qml.Identity
now accepts multiple wires. (#3049)
>>> id_op = qml.Identity([0, 1])
>>> id_op.matrix()
array([[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.]])
>>> id_op.sparse_matrix()
<4x4 sparse matrix of type '<class 'numpy.float64'>'
with 4 stored elements in Compressed Sparse Row format>
>>> id_op.eigvals()
array([1., 1., 1., 1.])
Added unitary_check
keyword argument to the constructor of the QubitUnitary
class which indicates whether the user wants to check for unitarity of the input matrix or not. Its default value is false
. (#3063)
Modified the representation of WireCut
by using qml.draw_mpl
. (#3067)
Improved the performance of qml.math.expand_matrix
function for dense and sparse matrices. (#3060) (#3064)
Added support for sums and products of operator classes with scalar tensors of any interface (NumPy, JAX, Tensorflow, PyTorch...). (#3149)
>>> s_prod = torch.tensor(4) * qml.RX(1.23, 0)
>>> s_prod
4*(RX(1.23, wires=[0]))
>>> s_prod.scalar
tensor(4)
Added overlapping_ops
property to the Composite
class to improve the performance of the eigvals
, diagonalizing_gates
and Prod.matrix
methods. (#3084)
Added the map_wires
method to the operators, which returns a copy of the operator with its wires changed according to the given wire map. (#3143)
>>> op = qml.Toffoli([0, 1, 2])
>>> wire_map = {0: 2, 2: 0}
>>> op.map_wires(wire_map=wire_map)
Toffoli(wires=[2, 1, 0])
Calling compute_matrix
and compute_sparse_matrix
of simple non-parametric operations is now faster and more memory-efficient with the addition of caching. (#3134)
Added details to the output of Exp.label()
. (#3126)
qml.math.unwrap
no longer creates ragged arrays. Lists remain lists. (#3163)
New null.qubit
device. The null.qubit
performs no operations or memory allocations. (#2589)
default.qubit
favours decomposition and avoids matrix construction for QFT
and GroverOperator
at larger qubit numbers. (#3193)
qml.ControlledQubitUnitary
now has a control_values
property. (#3206)
Added a new qml.tape.QuantumScript
class that contains all the non-queuing behavior of QuantumTape
. Now, QuantumTape
inherits from QuantumScript
as well as AnnotatedQueue
. (#3097)
Extended the qml.equal
function to MeasurementProcesses (#3189)
qml.drawer.draw.draw_mpl
now accepts a style
kwarg to select a style for plotting, rather than calling qml.drawer.use_style(style)
before plotting. Setting a style for draw_mpl
does not change the global configuration for matplotlib plotting. If no style
is passed, the function defaults to plotting with the black_white
style. (#3247)
QuantumTape._par_info
is now a list of dictionaries, instead of a dictionary whose keys are integers starting from zero. (#3185)
QueuingContext
has been renamed to QueuingManager
. (#3061)
Deprecation patches for the return types enum's location and qml.utils.expand
are removed. (#3092)
_multi_dispatch
functionality has been moved inside the get_interface
function. This function can now be called with one or multiple tensors as arguments. (#3136)
>>> torch_scalar = torch.tensor(1)
>>> torch_tensor = torch.Tensor([2, 3, 4])
>>> numpy_tensor = np.array([5, 6, 7])
>>> qml.math.get_interface(torch_scalar)
'torch'
>>> qml.math.get_interface(numpy_tensor)
'numpy'
_multi_dispatch
previously had only one argument which contained a list of the tensors to be dispatched:
>>> qml.math._multi_dispatch([torch_scalar, torch_tensor, numpy_tensor])
'torch'
To differentiate whether the user wants to get the interface of a single tensor or multiple tensors, get_interface
now accepts a different argument per tensor to be dispatched:
>>> qml.math.get_interface(*[torch_scalar, torch_tensor, numpy_tensor])
'torch'
>>> qml.math.get_interface(torch_scalar, torch_tensor, numpy_tensor)
'torch'
Operator.compute_terms
is removed. On a specific instance of an operator, op.terms()
can be used instead. There is no longer a static method for this. (#3215)
QueuingManager.safe_update_info
and AnnotatedQueue.safe_update_info
are deprecated. Instead, update_info
no longer raises errors if the object isn't in the queue. (#3085)
qml.tape.stop_recording
and QuantumTape.stop_recording
have been moved to qml.QueuingManager.stop_recording
. The old functions will still be available until v0.29. (#3068)
qml.tape.get_active_tape
has been deprecated. Use qml.QueuingManager.active_context()
instead. (#3068)
Operator.compute_terms
has been removed. On a specific instance of an operator, use op.terms()
instead. There is no longer a static method for this. (#3215)
qml.tape.QuantumTape.inv()
has been deprecated. Use qml.tape.QuantumTape.adjoint
instead. (#3237)
qml.transforms.qcut.remap_tape_wires
has been deprecated. Use qml.map_wires
instead. (#3186)
The grouping module qml.grouping
has been deprecated. Use qml.pauli
or qml.pauli.grouping
instead. The module will still be available until v0.28. (#3262)
The code block in the usage details of the UCCSD template has been updated. (#3140)
Added a "Deprecations" page to the developer documentation. (#3093)
The example of the qml.FlipSign
template has been updated. (#3219)
qml.SparseHamiltonian
now validates the size of the input matrix. (#3278)
Users no longer see unintuitive errors when inputing sequences to qml.Hermitian
. (#3181)
The evaluation of QNodes that return either vn_entropy
or mutual_info
raises an informative error message when using devices that define a vector of shots. (#3180)
Fixed a bug that made qml.AmplitudeEmbedding
incompatible with JITting. (#3166)
Fixed the qml.transforms.transpile
transform to work correctly for all two-qubit operations. (#3104)
Fixed a bug with the control values of a controlled version of a ControlledQubitUnitary
. (#3119)
Fixed a bug where qml.math.fidelity(non_trainable_state, trainable_state)
failed unexpectedly. (#3160)
Fixed a bug where qml.QueuingManager.stop_recording
did not clean up if yielded code raises an exception. (#3182)
Returning qml.sample()
or qml.counts()
with other measurements of non-commuting observables now raises a QuantumFunctionError (e.g., return qml.expval(PauliX(wires=0)), qml.sample()
now raises an error). (#2924)
Fixed a bug where op.eigvals()
would return an incorrect result if the operator was a non-hermitian composite operator. (#3204)
Fixed a bug where qml.BasisStatePreparation
and qml.BasisEmbedding
were not jit-compilable with JAX. (#3239)
Fixed a bug where qml.MottonenStatePreparation
was not jit-compilable with JAX. (#3260)
Fixed a bug where qml.expval(qml.Hamiltonian())
would not raise an error if the Hamiltonian involved some wires that are not present on the device. (#3266)
Fixed a bug where qml.tape.QuantumTape.shape()
did not account for the batch dimension of the tape (#3269)
This release contains contributions from (in alphabetical order):
Kamal Mohamed Ali, Guillermo Alonso-Linaje, Juan Miguel Arrazola, Utkarsh Azad, Thomas Bromley, Albert Mitjans Coma, Isaac De Vlugt, Olivia Di Matteo, Amintor Dusko, Lillian M. A. Frederiksen, Diego Guala, Josh Izaac, Soran Jahangiri, Edward Jiang, Korbinian Kottmann, Christina Lee, Romain Moyard, Lee J. O'Riordan, Mudit Pandey, Matthew Silverman, Jay Soni, Antal Száva, David Wierichs.
Published by josh146 about 2 years ago
A minor postfix release to update the documentation styling.
Published by rmoyard about 2 years ago
PennyLane now provides built-in support for implementing the classical-shadows measurement protocol. (#2820) (#2821) (#2871) (#2968) (#2959) (#2968)
The classical-shadow measurement protocol is described in detail in the paper Predicting Many Properties of a Quantum System from Very Few Measurements. As part of the support for classical shadows in this release, two new finite-shot and fully-differentiable measurements are available:
QNodes returning the new measurement qml.classical_shadow()
will return two entities; bits
(0 or 1 if the 1 or -1 eigenvalue is sampled, respectively) and recipes
(the randomized Pauli measurements that are performed for each qubit, labelled by integer):
dev = qml.device("default.qubit", wires=2, shots=3)
@qml.qnode(dev)
def circuit():
qml.Hadamard(wires=0)
qml.CNOT(wires=[0, 1])
return qml.classical_shadow(wires=[0, 1])
>>> bits, recipes = circuit()
>>> bits
tensor([[0, 0],
[1, 0],
[0, 1]], dtype=uint8, requires_grad=True)
>>> recipes
tensor([[2, 2],
[0, 2],
[0, 2]], dtype=uint8, requires_grad=True)
QNodes returning qml.shadow_expval()
yield the expectation value estimation using classical shadows:
dev = qml.device("default.qubit", wires=range(2), shots=10000)
@qml.qnode(dev)
def circuit(x, H):
qml.Hadamard(0)
qml.CNOT((0,1))
qml.RX(x, wires=0)
return qml.shadow_expval(H)
x = np.array(0.5, requires_grad=True)
H = qml.Hamiltonian(
[1., 1.],
[qml.PauliZ(0) @ qml.PauliZ(1), qml.PauliX(0) @ qml.PauliX(1)]
)
>>> circuit(x, H)
tensor(1.8486, requires_grad=True)
>>> qml.grad(circuit)(x, H)
-0.4797000000000001
Fully-differentiable QNode transforms for both new classical-shadows measurements are also available via qml.shadows.shadow_state
and qml.shadows.shadow_expval
, respectively.
For convenient post-processing, we've also added the ability to calculate general Renyi entropies by way of the ClassicalShadow
class' entropy
method, which requires the wires of the subsystem of interest and the Renyi entropy order:
>>> shadow = qml.ClassicalShadow(bits, recipes)
>>> vN_entropy = shadow.entropy(wires=[0, 1], alpha=1)
An entirely new framework for quantum computing is now simulatable with the addition of qutrit functionalities. (#2699) (#2781) (#2782) (#2783) (#2784) (#2841) (#2843)
Qutrits are like qubits, but instead live in a three-dimensional Hilbert space; they are not binary degrees of freedom, they are tertiary. The advent of qutrits allows for all sorts of interesting theoretical, practical, and algorithmic capabilities that have yet to be discovered.
To facilitate qutrit circuits requires a new device: default.qutrit
. The default.qutrit
device is a Python-based simulator, akin to default.qubit
, and is defined as per usual:
>>> dev = qml.device("default.qutrit", wires=1)
The following operations are supported on default.qutrit
devices:
qml.TShift
, and the ternary clock operator, qml.TClock
, as defined in this paper by Yeh et al. (2022),qml.TAdd
and qml.TSWAP
operations which are the qutrit analogs of the CNOT and SWAP operations, respectively.qml.QutritUnitary
.qml.state
and qml.probs
measurements.qml.THermitian
.A comprehensive example of these features is given below:
dev = qml.device("default.qutrit", wires=1)
U = np.array([
[1, 1, 1],
[1, 1, 1],
[1, 1, 1]
]
) / np.sqrt(3)
obs = np.array([
[1, 1, 0],
[1, -1, 0],
[0, 0, np.sqrt(2)]
]
) / np.sqrt(2)
@qml.qnode(dev)
def qutrit_state(U, obs):
qml.TShift(0)
qml.TClock(0)
qml.QutritUnitary(U, wires=0)
return qml.state()
@qml.qnode(dev)
def qutrit_expval(U, obs):
qml.TShift(0)
qml.TClock(0)
qml.QutritUnitary(U, wires=0)
return qml.expval(qml.THermitian(obs, wires=0))
>>> qutrit_state(U, obs)
tensor([-0.28867513+0.5j, -0.28867513+0.5j, -0.28867513+0.5j], requires_grad=True)
>>> qutrit_expval(U, obs)
tensor(0.80473785, requires_grad=True)
We will continue to add more and more support for qutrits in future releases.
The qml.simplify()
function has several intuitive improvements with this release. (#2978) (#2982) (#2922) (#3012)
qml.simplify
can now perform the following:
Here is an example of qml.simplify
in action with parameterized rotation gates. In this case, the angles of rotation are simplified to be modulo $4\pi$.
>>> op1 = qml.RX(30.0, wires=0)
>>> qml.simplify(op1)
RX(4.867258771281655, wires=[0])
>>> op2 = qml.RX(4 * np.pi, wires=0)
>>> qml.simplify(op2)
Identity(wires=[0])
All of these simplification features can be applied directly to quantum functions, QNodes, and tapes via decorating with @qml.simplify
, as well:
dev = qml.device("default.qubit", wires=2)
@qml.simplify
@qml.qnode(dev)
def circuit():
qml.adjoint(qml.prod(qml.RX(1, 0) ** 1, qml.RY(1, 0), qml.RZ(1, 0)))
return qml.probs(wires=0)
>>> circuit()
>>> list(circuit.tape)
[RZ(11.566370614359172, wires=[0]) @ RY(11.566370614359172, wires=[0]) @ RX(11.566370614359172, wires=[0]),
probs(wires=[0])]
A new optimizer called qml.QNSPSAOptimizer
is available that implements the quantum natural simultaneous perturbation stochastic approximation (QNSPSA) method based on Simultaneous Perturbation Stochastic Approximation of the Quantum Fisher Information. (#2818)
qml.QNSPSAOptimizer
is a second-order SPSA algorithm, which combines the convergence power of the quantum-aware Quantum Natural Gradient (QNG) optimization method with the reduced quantum evaluations of SPSA methods.
While the QNSPSA optimizer requires additional circuit executions (10 executions per step) compared to standard SPSA optimization (3 executions per step), these additional evaluations are used to provide a stochastic estimation of a second-order metric tensor, which often helps the optimizer to achieve faster convergence.
Use qml.QNSPSAOptimizer
like you would any other optimizer:
max_iterations = 50
opt = qml.QNSPSAOptimizer()
for _ in range(max_iterations):
params, cost = opt.step_and_cost(cost, params)
Check out our demo on the QNSPSA optimizer for more information.
Operator methods for exponentiation and raising to a power have been added. (#2799) (#3029)
The qml.exp
function can be used to create observables or generic rotation gates:
>>> x = 1.234
>>> t = qml.PauliX(0) @ qml.PauliX(1) + qml.PauliY(0) @ qml.PauliY(1)
>>> isingxy = qml.exp(t, 0.25j * x)
>>> isingxy.matrix()
array([[1. +0.j , 0. +0.j ,
1. +0.j , 0. +0.j ],
[0. +0.j , 0.8156179+0.j ,
1. +0.57859091j, 0. +0.j ],
[0. +0.j , 0. +0.57859091j,
0.8156179+0.j , 0. +0.j ],
[0. +0.j , 0. +0.j ,
1. +0.j , 1. +0.j ]])
The qml.pow
function raises a given operator to a power:
>>> op = qml.pow(qml.PauliX(0), 2)
>>> op.matrix()
array([[1, 0], [0, 1]])
An operator called qml.PSWAP
is now available. (#2667)
The qml.PSWAP
gate -- or phase-SWAP gate -- was previously available within the PennyLane-Braket plugin only. Enjoy it natively in PennyLane with v0.26.
Check whether or not an operator is hermitian or unitary with qml.is_hermitian
and qml.is_unitary
. (#2960)
>>> op1 = qml.PauliX(wires=0)
>>> qml.is_hermitian(op1)
True
>>> op2 = qml.PauliX(0) + qml.RX(np.pi/3, 0)
>>> qml.is_unitary(op2)
False
Embedding templates now support parameter broadcasting. (#2810)
Embedding templates like AmplitudeEmbedding
or IQPEmbedding
now support parameter broadcasting with a leading broadcasting dimension in their variational parameters. AmplitudeEmbedding
, for example, would usually use a one-dimensional input vector of features. With broadcasting, we can now compute
>>> features = np.array([
... [0.5, 0.5, 0., 0., 0.5, 0., 0.5, 0.],
... [1., 0., 0., 0., 0., 0., 0., 0.],
... [0.5, 0.5, 0., 0., 0., 0., 0.5, 0.5],
... ])
>>> op = qml.AmplitudeEmbedding(features, wires=[1, 5, 2])
>>> op.batch_size
3
An exception is BasisEmbedding
, which is not broadcastable.
The qml.math.expand_matrix()
method now allows the sparse matrix representation of an operator to be extended to a larger hilbert space. (#2998)
>>> from scipy import sparse
>>> mat = sparse.csr_matrix([[0, 1], [1, 0]])
>>> qml.math.expand_matrix(mat, wires=[1], wire_order=[0,1]).toarray()
array([[0., 1., 0., 0.],
[1., 0., 0., 0.],
[0., 0., 0., 1.],
[0., 0., 1., 0.]])
qml.ctrl
now uses Controlled
instead of ControlledOperation
. The new Controlled
class wraps individual Operator
's instead of a tape. It provides improved representations and integration. (#2990)
qml.matrix
can now compute the matrix of tapes and QNodes that contain multiple broadcasted operations or non-broadcasted operations after broadcasted ones. (#3025)
A common scenario in which this becomes relevant is the decomposition of broadcasted operations: the decomposition in general will contain one or multiple broadcasted operations as well as operations with no or fixed parameters that are not broadcasted.
Lists of operators are now internally sorted by their respective wires while also taking into account their commutativity property.(#2995)
Some methods of the QuantumTape
class have been simplified and reordered to improve both readability and performance. (#2963)
The qml.qchem.molecular_hamiltonian
function is modified to support observable grouping. (#2997)
qml.ops.op_math.Controlled
now has basic decomposition functionality. (#2938)
Automatic circuit cutting has been improved by making better partition imbalance derivations. Now it is more likely to generate optimal cuts for larger circuits. (#2517)
By default, qml.counts
only returns the outcomes observed in sampling. Optionally, specifying qml.counts(all_outcomes=True)
will return a dictionary containing all possible outcomes. (#2889)
>>> dev = qml.device("default.qubit", wires=2, shots=1000)
>>>
>>> @qml.qnode(dev)
>>> def circuit():
... qml.Hadamard(wires=0)
... qml.CNOT(wires=[0, 1])
... return qml.counts(all_outcomes=True)
>>> result = circuit()
>>> result
{'00': 495, '01': 0, '10': 0, '11': 505}
Internal use of in-place inversion is eliminated in preparation for its deprecation. (#2965)
Controlled
operators now work with qml.is_commuting
. (#2994)
qml.prod
and qml.op_sum
now support the sparse_matrix()
method. (#3006)
>>> xy = qml.prod(qml.PauliX(1), qml.PauliY(1))
>>> op = qml.op_sum(xy, qml.Identity(0))
>>>
>>> sparse_mat = op.sparse_matrix(wire_order=[0,1])
>>> type(sparse_mat)
<class 'scipy.sparse.csr.csr_matrix'>
>>> sparse_mat.toarray()
[[1.+1.j 0.+0.j 0.+0.j 0.+0.j]
[0.+0.j 1.-1.j 0.+0.j 0.+0.j]
[0.+0.j 0.+0.j 1.+1.j 0.+0.j]
[0.+0.j 0.+0.j 0.+0.j 1.-1.j]]
Provided sparse_matrix()
support for single qubit observables. (#2964)
qml.Barrier
with only_visual=True
now simplifies via op.simplify()
to the identity operator or a product of identity operators.(#3016)
More accurate and intuitive outputs for printing some operators have been added. (#3013)
Results for the matrix of the sum or product of operators are stored in a more efficient manner. (#3022)
The computation of the (sparse) matrix for the sum or product of operators is now more efficient. (#3030)
When the factors of qml.prod
don't share any wires, the matrix and sparse matrix are computed using a kronecker product for improved efficiency. (#3040)
qml.grouping.is_pauli_word
now returns False
for operators that don't inherit from qml.Observable
instead of raising an error. (#3039)
Added functionality to iterate over operators created from qml.op_sum
and qml.prod
. (#3028)
>>> op = qml.op_sum(qml.PauliX(0), qml.PauliY(1), qml.PauliZ(2))
>>> len(op)
3
>>> op[1]
PauliY(wires=[1])
>>> [o.name for o in op]
['PauliX', 'PauliY', 'PauliZ']
In-place inversion is now deprecated. This includes op.inv()
and op.inverse=value
. Please use qml.adjoint
or qml.pow
instead. Support for these methods will remain till v0.28. (#2988)
Don't use:
>>> v1 = qml.PauliX(0).inv()
>>> v2 = qml.PauliX(0)
>>> v2.inverse = True
Instead use:
>>> qml.adjoint(qml.PauliX(0))
Adjoint(PauliX(wires=[0]))
>>> qml.pow(qml.PauliX(0), -1)
PauliX(wires=[0])**-1
>>> qml.pow(qml.PauliX(0), -1, lazy=False)
PauliX(wires=[0])
>>> qml.PauliX(0) ** -1
PauliX(wires=[0])**-1
qml.adjoint
takes the conjugate transpose of an operator, while qml.pow(op, -1)
indicates matrix inversion. For unitary operators, adjoint
will be more efficient than qml.pow(op, -1)
, even though they represent the same thing.
The supports_reversible_diff
device capability is unused and has been removed. (#2993)
Measuring an operator that might not be hermitian now raises a warning instead of an error. To definitively determine whether or not an operator is hermitian, use qml.is_hermitian
. (#2960)
The ControlledOperation
class has been removed. This was a developer-only class, so the change should not be evident to any users. It is replaced by Controlled
. (#2990)
The default execute
method for the QubitDevice
base class now calls self.statistics
with an additional keyword argument circuit
, which represents the quantum tape being executed. Any device that overrides statistics
should edit the signature of the method to include the new circuit
keyword argument. (#2820)
The expand_matrix()
has been moved from pennylane.operation
to pennylane.math.matrix_manipulation
. (#3008)
qml.grouping.utils.is_commuting
has been removed, and its Pauli word logic is now part of qml.is_commuting
. (#3033)
qml.is_commuting
has been moved from pennylane.transforms.commutation_dag
to pennylane.ops.functions
. (#2991)
Updated the Fourier transform docs to use circuit_spectrum
instead of spectrum
, which has been deprecated. (#3018)
Corrected the docstrings for diagonalizing gates for all relevant operations. The docstrings used to say that the diagonalizing gates implemented $U$, the unitary such that $O = U \Sigma U^{\dagger}$, where $O$ is the original observable and $\Sigma$ a diagonal matrix. However, the diagonalizing gates actually implement $U^{\dagger}$, since $\langle \psi | O | \psi \rangle = \langle \psi | U \Sigma U^{\dagger} | \psi \rangle$, making $U^{\dagger} | \psi \rangle$ the actual state being measured in the Z-basis. (#2981)
Fixed a bug with qml.ops.Exp
operators when the coefficient is autograd but the diagonalizing gates don't act on all wires. (#3057)
Fixed a bug where the tape transform single_qubit_fusion
computed wrong rotation angles for specific combinations of rotations. (#3024)
Jax gradients now work with a QNode when the quantum function was transformed by qml.simplify
. (#3017)
Operators that have num_wires = AnyWires
or num_wires = AnyWires
now raise an error, with certain exceptions, when instantiated with wires=[]
. (#2979)
Fixed a bug where printing qml.Hamiltonian
with complex coefficients raises TypeError
in some cases. (#3005)
Added a more descriptive error message when measuring non-commuting observables at the end of a circuit with probs
, samples
, counts
and allcounts
. (#3065)
This release contains contributions from (in alphabetical order):
Juan Miguel Arrazola, Utkarsh Azad, Tom Bromley, Olivia Di Matteo, Isaac De Vlugt, Yiheng Duan, Lillian Marie Austin Frederiksen, Josh Izaac, Soran Jahangiri, Edward Jiang, Ankit Khandelwal, Korbinian Kottmann, Meenu Kumari, Christina Lee, Albert Mitjans Coma, Romain Moyard, Rashid N H M, Zeyue Niu, Mudit Pandey, Matthew Silverman, Jay Soni, Antal Száva, Cody Wang, David Wierichs.
Published by rmoyard about 2 years ago
qml.math.array
and qml.math.eye
to preserve the Torch device used. (#2967)
This release contains contributions from (in alphabetical order):
Romain Moyard, Rashid N H M, Lee James O'Riordan, Antal Száva.