Source code for pennylane.templates.subroutines.qchem.all_singles_doubles

# Copyright 2018-2021 Xanadu Quantum Technologies Inc.

# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at

#     http://www.apache.org/licenses/LICENSE-2.0

# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
r"""
Contains the AllSinglesDoubles template.
"""

# pylint: disable=too-many-arguments,protected-access
import copy
from collections.abc import Sequence

import numpy as np

from pennylane import capture, math
from pennylane.control_flow import for_loop
from pennylane.decomposition import add_decomps, register_resources, resource_rep
from pennylane.operation import Operation
from pennylane.ops import BasisState, DoubleExcitation, SingleExcitation
from pennylane.typing import TensorLike
from pennylane.wires import Wires, WiresLike

has_jax = True
try:
    from jax import numpy as jnp
except (ModuleNotFoundError, ImportError) as import_error:  # pragma: no cover
    has_jax = False  # pragma: no cover


[docs] class AllSinglesDoubles(Operation): r"""Builds a quantum circuit to prepare correlated states of molecules by applying all :class:`~.pennylane.SingleExcitation` and :class:`~.pennylane.DoubleExcitation` operations to the initial Hartree-Fock state. The template initializes the :math:`n`-qubit system to encode the input Hartree-Fock state and applies the particle-conserving :class:`~.pennylane.SingleExcitation` and :class:`~.pennylane.DoubleExcitation` operations which are implemented as `Givens rotations <https://en.wikipedia.org/wiki/Givens_rotation>`_ that act on the subspace of two and four qubits, respectively. The total number of excitation gates and the indices of the qubits they act on are obtained using the :func:`~.excitations` function. For example, the quantum circuit for the case of two electrons and six qubits is sketched in the figure below: | .. figure:: ../../_static/templates/subroutines/all_singles_doubles.png :align: center :width: 70% :target: javascript:void(0); | In this case, we have four single and double excitations that preserve the total-spin projection of the Hartree-Fock state. The :class:`~.pennylane.SingleExcitation` gate :math:`G` act on the qubits ``[0, 2], [0, 4], [1, 3], [1, 5]`` as indicated by the squares, while the :class:`~.pennylane.DoubleExcitation` operation :math:`G^{(2)}` is applied to the qubits ``[0, 1, 2, 3], [0, 1, 2, 5], [0, 1, 2, 4], [0, 1, 4, 5]``. The resulting unitary conserves the number of particles and prepares the :math:`n`-qubit system in a superposition of the initial Hartree-Fock state and other states encoding multiply-excited configurations. Args: weights (TensorLike): size ``(len(singles) + len(doubles),)`` tensor containing the angles entering the :class:`~.pennylane.SingleExcitation` and :class:`~.pennylane.DoubleExcitation` operations, in that order wires (WiresLike): wires that the template acts on hf_state (Sequence[int]): Length ``len(wires)`` occupation-number vector representing the Hartree-Fock state. ``hf_state`` is used to initialize the wires. singles (Sequence[tuple[int, int]] | None): An optional sequence of lists with the indices of the two qubits the :class:`~.pennylane.SingleExcitation` operations act on doubles (Sequence[tuple[int, int, int, int]] | None): An optional sequence of lists with the indices of the four qubits the :class:`~.pennylane.DoubleExcitation` operations act on .. details:: :title: Usage Details Notice that: #. The number of wires has to be equal to the number of spin orbitals included in the active space. #. The single and double excitations can be generated with the function :func:`~.excitations`. See example below. An example of how to use this template is shown below: .. code-block:: python import pennylane as qml import numpy as np electrons = 2 qubits = 4 # Define the HF state hf_state = qml.qchem.hf_state(electrons, qubits) # Generate all single and double excitations singles, doubles = qml.qchem.excitations(electrons, qubits) # Define the device dev = qml.device('default.qubit', wires=qubits) wires = range(qubits) @qml.qnode(dev) def circuit(weights, hf_state, singles, doubles): qml.templates.AllSinglesDoubles(weights, wires, hf_state, singles, doubles) return qml.expval(qml.Z(0)) # Evaluate the QNode for a given set of parameters params = np.random.normal(0, np.pi, len(singles) + len(doubles)) circuit(params, hf_state, singles=singles, doubles=doubles) """ grad_method = None resource_keys = {"num_singles", "num_doubles", "num_wires"} def __init__( self, weights: TensorLike, wires: WiresLike, hf_state: Sequence[int], singles: Sequence[tuple[int, int]] | None = None, doubles: Sequence[tuple[int, int, int, int]] | None = None, id=None, ): wires = Wires(wires) if len(wires) < 2: raise ValueError( f"The number of qubits (wires) can not be less than 2; got len(wires) = {len(wires)}" ) if doubles is not None: if any(len(d_wires) != 4 for d_wires in doubles): raise ValueError("Expected all entries of 'doubles' to be of size 4.") doubles = () if doubles is None else tuple(tuple(d) for d in doubles) if singles is not None: if any(len(s_wires) != 2 for s_wires in singles): raise ValueError("Expected all entries of 'singles' to be of size 2.") singles = () if singles is None else tuple(tuple(s) for s in singles) weights_shape = math.shape(weights) exp_shape = self.shape(singles, doubles) if weights_shape != exp_shape: raise ValueError(f"'weights' tensor must be of shape {exp_shape}; got {weights_shape}.") if len(hf_state) != len(wires): raise ValueError( f"Expected length of 'hf_state' to match number of wires ({len(wires)})." ) if (hf_dtype := hf_state[0].dtype) != np.dtype("int"): raise ValueError(f"Elements of 'hf_state' must be integers, got {hf_dtype}.") hf_state = tuple(hf_state) self._hyperparameters = { "hf_state": hf_state, "singles": singles, "doubles": doubles, } super().__init__(weights, wires=wires, id=id) @classmethod def _primitive_bind_call( cls, weights, wires, hf_state, singles=None, doubles=None, id=None ): # pylint: disable=arguments-differ singles = math.array(singles) if singles is not None else math.array(((),)) doubles = math.array(doubles) if doubles is not None else math.array(((),)) wires = math.array(wires) hf_state = math.array(hf_state) weights = math.array(weights) return cls._primitive.bind(weights, wires, hf_state, singles, doubles, id=id) @property def resource_params(self) -> dict: return { "num_singles": len(self.hyperparameters["singles"]), "num_doubles": len(self.hyperparameters["doubles"]), "num_wires": len(self.wires), }
[docs] def map_wires(self, wire_map: dict): new_op = copy.deepcopy(self) new_op._wires = Wires([wire_map.get(wire, wire) for wire in self.wires]) for key in ["singles", "doubles"]: new_op._hyperparameters[key] = tuple( tuple(wire_map[w] for w in wires) for wires in new_op._hyperparameters[key] ) return new_op
@property def num_params(self): return 1
[docs] @staticmethod def compute_decomposition( weights, wires, hf_state, singles, doubles ): # pylint: disable=arguments-differ r"""Representation of the operator as a product of other operators. .. math:: O = O_1 O_2 \dots O_n. .. seealso:: :meth:`~.AllSinglesDoubles.decomposition`. Args: weights (tensor_like): size ``(len(singles) + len(doubles),)`` tensor containing the angles entering the :class:`~.pennylane.SingleExcitation` and :class:`~.pennylane.DoubleExcitation` operations, in that order wires (Any or Iterable[Any]): wires that the BasisState operator acts on hf_state (array[int]): Length ``len(wires)`` occupation-number vector representing the Hartree-Fock state. ``hf_state`` is used to initialize the wires. singles (Sequence[Sequence]): sequence of lists with the indices of the two qubits the :class:`~.pennylane.SingleExcitation` operations act on doubles (Sequence[Sequence]): sequence of lists with the indices of the four qubits the :class:`~.pennylane.DoubleExcitation` operations act on Returns: list[.Operator]: decomposition of the operator """ op_list = [] op_list.append(BasisState(hf_state, wires=wires)) for i, d_wires in enumerate(doubles): op_list.append(DoubleExcitation(weights[len(singles) + i], wires=d_wires)) for j, s_wires in enumerate(singles): op_list.append(SingleExcitation(weights[j], wires=s_wires)) return op_list
[docs] @staticmethod def shape(singles, doubles): r"""Returns the expected shape of the tensor that contains the circuit parameters. Args: singles (Sequence[Sequence]): sequence of lists with the indices of the two qubits the :class:`~.pennylane.SingleExcitation` operations act on doubles (Sequence[Sequence]): sequence of lists with the indices of the four qubits the :class:`~.pennylane.DoubleExcitation` operations act on Returns: tuple(int): shape of the tensor containing the circuit parameters """ if (singles is None or len(singles) == 0) and (doubles is None or len(doubles) == 0): raise ValueError( f"'singles' and 'doubles' lists can not be both empty;" f" got singles = {singles}, doubles = {doubles}" ) if singles is None or len(singles) == 0: shape_ = (len(doubles),) elif doubles is None or len(doubles) == 0: shape_ = (len(singles),) else: shape_ = (len(singles) + len(doubles),) return shape_
if AllSinglesDoubles._primitive is not None: @AllSinglesDoubles._primitive.def_impl def _(*args, **kwargs): # pylint: disable=unused-argument # need to convert array values into integers # for plxpr, all wires must be integers # could be abstract when using tracing evaluation in interpreter wires = tuple(w if math.is_abstract(w) else int(w) for w in args[1]) return type.__call__(AllSinglesDoubles, args[0], wires, args[2], args[3], args[4]) def _all_singles_doubles_resouces(num_singles, num_doubles, num_wires): return { resource_rep(BasisState, num_wires=num_wires): 1, resource_rep(DoubleExcitation): num_doubles, resource_rep(SingleExcitation): num_singles, } @register_resources(_all_singles_doubles_resouces) def _all_singles_doubles_decomposition(weights, wires, hf_state, singles, doubles): BasisState(hf_state, wires=wires) if has_jax and capture.enabled(): weights, doubles, singles = jnp.array(weights), jnp.array(doubles), jnp.array(singles) @for_loop(len(doubles)) def doubles_loop(i): DoubleExcitation(weights[len(singles) + i], wires=doubles[i]) doubles_loop() # pylint: disable=no-value-for-parameter @for_loop(len(singles)) def singles_loop(j): SingleExcitation(weights[j], wires=singles[j]) singles_loop() # pylint: disable=no-value-for-parameter add_decomps(AllSinglesDoubles, _all_singles_doubles_decomposition)