"""
\********************************************************************************
* Copyright (c) 2023 the Qrisp authors
*
* This program and the accompanying materials are made available under the
* terms of the Eclipse Public License 2.0 which is available at
* http://www.eclipse.org/legal/epl-2.0.
*
* This Source Code may also be made available under the following Secondary
* Licenses when the conditions for such availability set forth in the Eclipse
* Public License, v. 2.0 are satisfied: GNU General Public License, version 2
* with the GNU Classpath Exception which is
* available at https://www.gnu.org/software/classpath/license.html.
*
* SPDX-License-Identifier: EPL-2.0 OR GPL-2.0 WITH Classpath-exception-2.0
********************************************************************************/
"""
import copy
import weakref
import matplotlib.pyplot as plt
import numpy as np
from qrisp.core.compilation import qompiler
[docs]class QuantumVariable:
"""
The QuantumVariable is the quantum equivalent of a regular variable in classical
programming languages. All :ref:`quantum types <QuantumTypes>` inherit from this
class. The QuantumVariable allows many automizations and quality of life
improvements such as hidden qubit management, de/encoding to human readable labels
or typing.
Each QuantumVariable is registered in a :ref:`QuantumSession`. It can be accessed
using the ``.qs`` attribute:
>>> from qrisp import QuantumVariable
>>> example_qv = QuantumVariable(3)
>>> quantum_session = example_qv.qs
The qubits of the QuantumVariable are stored as a list in the ``.reg`` attribute
>>> qubits = example_qv.reg
To quickly access the qubits of a given variable, we use the [ ] operator:
>>> qubit_2 = example_qv[2]
We can find out about the amount of qubits in the QuantumVariable with the ``.size``
attribute
>>> example_qv.size
3
**Naming**
QuantumVariables can be given names to identify them independently of their naming
as Python objects.
>>> example_qv_2 = QuantumVariable(3, name = "alice")
>>> example_qv_2.name
'alice'
If not explicitely specified during construction, a name is determined
automatically. Qrisp will try to infer the name of the Python variable and if that
fails, a generic name is given.
>>> example_qv.name
'example_qv'
In order to keep the generated quantum circuits comprehensive, the qubits are named
after their containing QuantumVariable with an extra number, which indicates their
index.
>>> from qrisp import cx
>>> cx(example_qv, example_qv_2)
>>> print(example_qv.qs)
::
QuantumCircuit:
--------------
example_qv.0: ──■────────────
│
example_qv.1: ──┼────■───────
│ │
example_qv.2: ──┼────┼────■──
┌─┴─┐ │ │
alice.0: ┤ X ├──┼────┼──
└───┘┌─┴─┐ │
alice.1: ─────┤ X ├──┼──
└───┘┌─┴─┐
alice.2: ──────────┤ X ├
└───┘
Live QuantumVariables:
---------------------
QuantumVariable example_qv
QuantumVariable alice
QuantumSessions can only contain uniquely named QuantumVariables. If two
QuantumSessions are :ref:`merged <SessionMerging>` containing identically named
QuantumVariables, the more recently created QuantumVariable will be renamed:
::
from qrisp import QuantumFloat
s = QuantumFloat(5)
for i in range(4):
temp = QuantumFloat(4)
temp[:] = 2**i
s += temp
>>> print(s.qs)
::
QuantumCircuit:
--------------
┌───────────┐┌───────────┐┌───────────┐┌───────────┐
s.0: ─────┤0 ├┤0 ├┤0 ├┤0 ├
│ ││ ││ ││ │
s.1: ─────┤1 ├┤1 ├┤1 ├┤1 ├
│ ││ ││ ││ │
s.2: ─────┤2 ├┤2 ├┤2 ├┤2 ├
│ ││ ││ ││ │
s.3: ─────┤3 ├┤3 ├┤3 ├┤3 ├
│ ││ ││ ││ │
s.4: ─────┤4 __iadd__ ├┤4 ├┤4 ├┤4 ├
┌───┐│ ││ ││ ││ │
temp.0: ┤ X ├┤5 ├┤ ├┤ ├┤ ├
└───┘│ ││ ││ ││ │
temp.1: ─────┤6 ├┤ __iadd__ ├┤ ├┤ ├
│ ││ ││ ││ │
temp.2: ─────┤7 ├┤ ├┤ ├┤ ├
│ ││ ││ ││ │
temp.3: ─────┤8 ├┤ ├┤ __iadd__ ├┤ ├
└───────────┘│ ││ ││ │
temp_1.0: ──────────────────┤5 ├┤ ├┤ ├
┌───┐ │ ││ ││ │
temp_1.1: ┤ X ├─────────────┤6 ├┤ ├┤ __iadd__ ├
└───┘ │ ││ ││ │
temp_1.2: ──────────────────┤7 ├┤ ├┤ ├
│ ││ ││ │
temp_1.3: ──────────────────┤8 ├┤ ├┤ ├
└───────────┘│ ││ │
temp_2.0: ───────────────────────────────┤5 ├┤ ├
│ ││ │
temp_2.1: ───────────────────────────────┤6 ├┤ ├
┌───┐ │ ││ │
temp_2.2: ┤ X ├──────────────────────────┤7 ├┤ ├
└───┘ │ ││ │
temp_2.3: ───────────────────────────────┤8 ├┤ ├
└───────────┘│ │
temp_3.0: ────────────────────────────────────────────┤5 ├
│ │
temp_3.1: ────────────────────────────────────────────┤6 ├
│ │
temp_3.2: ────────────────────────────────────────────┤7 ├
┌───┐ │ │
temp_3.3: ┤ X ├───────────────────────────────────────┤8 ├
└───┘ └───────────┘
Live QuantumVariables:
---------------------
QuantumFloat s
QuantumFloat temp
QuantumFloat temp_1
QuantumFloat temp_2
QuantumFloat temp_3
Renaming does not happen for names given through the ``name`` keyword, unless the
name ends with a ``*``.
>>> example_qv_3 = QuantumVariable(3, name = "alice")
>>> cx(example_qv, example_qv_3)
Exception: Tried to merge QuantumSession containing identically named
QuantumVariables
>>> example_qv_4 = QuantumVariable(3, name = "alice*")
>>> cx(example_qv, example_qv_4)
>>> example_qv_4.name
'alice_1'
Examples
--------
Writing a function that brings an arbitrary QuantumVariable into a GHZ state
::
from qrisp import QuantumVariable, h, cx
def GHZ(qv):
h(qv[0])
for i in range(1, qv.size):
cx(qv[0], qv[i])
Evaluation:
>>> qv = QuantumVariable(5)
>>> GHZ(qv)
>>> print(qv)
{'00000': 0.5, '11111': 0.5}
"""
live_qvs = []
creation_counter = np.zeros(1)
name_tracker = {}
[docs] def __init__(self, size, qs=None, name=None):
r"""
Constructs a QuantumVariable - possibly with a given name or in a given
QuantumSession.
Parameters
----------
size : int
The amount of qubits this QuantumVariable contains.
qs : QuantumSession, optional
A QuantumSession object, where the QuantumVariable is supposed to be
registered. The default is None.
name : string, optional
A name which uniquely identifies the QuantumVariable. If ended with a
\*, name is allowed to be updated if a naming collision arises. By default,
Qrisp will try to infer the name of the Python variable - otherwise a
generic name is given.
"""
if size < 0:
raise Exception(
f"Tried to create QuantumVariable with invalid qubit amount {size}"
)
# Store quantum session
from qrisp.core import QuantumSession, merge_sessions
if qs is not None:
self.qs = qs
else:
self.qs = QuantumSession()
self.size = int(size)
self.user_given_name = False
# If name is given, register variable in session manager
if name is not None:
self.user_given_name = True
if name[-1] == "*":
name = name[:-1]
self.user_given_name = False
try:
self.name = name
self.qs.register_qv(self)
except RuntimeError:
i = int(self.creation_counter)
while True:
try:
self.name = name + "_" + str(i)
self.qs.register_qv(self)
except RuntimeError:
i += 1
continue
break
else:
self.name = name
self.qs.register_qv(self)
# Otherwise try to infer from code inspection
else:
from qrisp.misc import find_calling_line
if type(self) is QuantumVariable:
line = find_calling_line(1)
else:
line = find_calling_line(2)
split_line = line.split("=")
name_found = False
if len(split_line) >= 2:
python_var_name = split_line[0]
if split_line[1].replace(" ", "")[:7] == "Quantum":
python_var_name = python_var_name.split(" ")[0]
python_var_name = python_var_name.split(" ")[-1]
# name = self.get_unique_name(python_var_name)
name = python_var_name
self.name = name
self.qs.register_qv(self)
name_found = True
# If this didn't work, generate a generic, unique name
if not name_found:
while True:
try:
self.name = self.get_unique_name()
self.qs.register_qv(self)
break
except RuntimeError:
pass
import weakref
# This attribute tracks the created QuantumVariables for the
# auto_uncompute decorator
# We use weak references as some qrisp modules rely on reference counting
QuantumVariable.live_qvs.append(weakref.ref(self))
self.creation_time = int(self.creation_counter[0])
self.creation_counter += 1
def __or__(self, other):
from qrisp import mcx, x, cx
if len(self) > len(other):
or_res = self.duplicate()
else:
or_res = other.duplicate()
for i in range(min(len(self), len(other))):
mcx([self[i], other[i]], or_res[i], ctrl_state = 0)
x(or_res[i])
for i in range(min(len(self), len(other)), len(self)):
cx(self[i], or_res[i])
for i in range(min(len(self), len(other)), len(other)):
cx(other[i], or_res[i])
return or_res
def __and__(self, other):
from qrisp import mcx
if len(self) > len(other):
and_res = self.duplicate()
else:
and_res = other.duplicate()
for i in range(min(len(self), len(other))):
mcx([self[i], other[i]], and_res[i])
return and_res
def __xor__(self, other):
from qrisp import cx
if len(self) > len(other):
and_res = self.duplicate()
else:
and_res = other.duplicate()
for i in range(min(len(self), len(other))):
cx(self[i], and_res[i])
cx(other[i], and_res[i])
return and_res
[docs] def delete(self, verify=False, recompute=False):
r"""
This method is for deleting a QuantumVariable and thus freeing up and resetting
the used qubits.
Note that this method has a different function than the destructor. Calling this
method will tell the QuantumSession to mark the used qubits as free and apply a
reset gate.
If set to True, the keyword verify will cause a simulation to check, wether the
deleted qubits are in the $\ket{0}$ state prior to resetting. This is helpfull
during debugging, as it indicates wether the uncomputation of this
QuantumVariable was successfull.
After deletion, the QuantumVariable object is basically unchanged but an error
will be raised if further operations on the deleted qubits are attempted.
Parameters
----------
verify : bool, optional
If this bool is set to True, Qrisp will verify that the deleted qubits are
indeed in the $\ket{0}$ state. The default is ``False``.
recompute : bool, optional
If set to ``True``, this QuantumVariable can be recomputed if it is required
for the uncomputation of another QuantumVariable. For more information on
the (dis)advantages check :ref:`recomputation <recomputation>`. The default
is ``False``.
Raises
------
Exception
Tried to delete qubits not in \|0> state.
Examples
--------
We create a QuantumVariable, execute some gates and try to delete with
verify = True
>>> from qrisp import QuantumVariable, x, h
>>> qv = QuantumVariable(2)
>>> x(qv[0])
>>> h(qv[1])
>>> qv.delete(verify = True)
Exception: Tried to delete qubits not in |0> state.
We now (manually) uncompute the gates
>>> x(qv[0])
>>> h(qv[1])
>>> qv.delete(verify = True)
>>> qv.is_deleted()
True
>>> x(qv[0])
Exception: Tried to perform operation x on unallocated qubit qv_1.0.
"""
if self.is_deleted():
return
self.qs.delete_qv(self, verify)
i = 0
while i < len(QuantumVariable.live_qvs):
if QuantumVariable.live_qvs[i]() is None:
QuantumVariable.live_qvs.pop(i)
continue
if QuantumVariable.live_qvs[i]().name == self.name:
QuantumVariable.live_qvs.pop(i)
break
i += 1
if recompute:
for qb in self.reg:
qb.recompute = True
def is_deleted(self):
for qb in self.reg:
if not qb.allocated:
return True
else:
return False
[docs] def duplicate(self, name=None, qs=None, init=False):
r"""
Duplicates the QuantumVariable in the sense that a new QuantumVariable is
created with same type and parameters but initialized in the $\ket{0}$ state.
Parameters
----------
name : string, optional
A unique name to identify that QuantumVariable. If not given, a name will be
generated.
qs : QuantumSession, optional
A QuantumSession, where the result should be registered. If not given, a new
QuantumSession will be generated.
init : bool, optional
If set to True, the :meth:`init_from <qrisp.QuantumVariable.init_from>`
method of the result will be called on self. The default is False.
Returns
-------
duplicate : Type of self
The duplicated QuantumVariable.
Examples
--------
We create a QuantumFloat and duplicate:
>>> from qrisp import QuantumFloat
>>> qf_0 = QuantumFloat(4, signed = False)
>>> qf_1 = qf_0.duplicate()
>>> type(qf_1)
qrisp.qtypes.quantum_float.QuantumFloat
>>> qf_1.size
4
"""
duplicate = copy.copy(self)
from qrisp.core import QuantumSession
new_qs = QuantumSession()
# Register duplicate variable in session manager
if name is not None:
if name[-1] == "*":
self.user_given_name = False
name = name[:-1]
else:
duplicate.user_given_name = True
duplicate.name = name
new_qs.register_qv(duplicate)
else:
duplicate.user_given_name = False
try:
duplicate.name = self.name + "_dupl"
new_qs.register_qv(duplicate)
except NameError:
i = 0
while True:
try:
duplicate.name = self.name + "_dupl" + str(i)
new_qs.register_qv(duplicate)
break
except NameError:
pass
i += 1
from qrisp import merge
duplicate.qs = new_qs
# This attribute tracks the created QuantumVariables for the
# auto_uncompute decorator
# We use weak references as some qrisp modules rely on reference counting
QuantumVariable.live_qvs.append(weakref.ref(duplicate))
duplicate.creation_time = int(self.creation_counter[0])
duplicate.creation_counter += 1
if qs is not None:
merge(qs, new_qs)
if init:
duplicate.init_from(self)
return duplicate
[docs] def decoder(self, i):
"""
The decoder method specifies how a QuantumVariable turns the outcomes of
measurements into human-readable values. It recieves an integer ``i`` and
returns a human-readable value.
This method is supposed to be overloaded when defining new
:ref:`quantum types <QuantumTypes>`.
Parameters
----------
i : int
Integer representing the outcome of a measurement of the qubits of this
QuantumVariable.
Returns
-------
A human-readable value. Has to be hashable.
Examples
--------
We create a QuantumFloat and inspect its decoder:
>>> from qrisp import QuantumFloat
>>> qf = QuantumFloat(3, -1, signed = False)
>>> print(qf.decoder(1))
0.5
This implies that if the 3 qubits of this QuantumFloat are measured in state
001, this outcome corresponds to the value 0.5.
"""
from qrisp.misc import bin_rep
return bin_rep(i, self.size)[::-1]
[docs] def encoder(self, value):
"""
The encoder reverses the decoder, it turns human-readable values into integers.
If not overloaded, the encoder will perform a linear search on decoder inputs to
match the given value.
Parameters
----------
label :
A human-readable value.
Raises
------
Exception
Unknown input value.
Returns
-------
i : int
The integer encoding the given value.
Examples
--------
We create a QuantumChar and inspect it's encoder:
>>> from qrisp import QuantumChar
>>> q_ch = QuantumChar()
>>> print(q_ch.encoder("f"))
5
This implies that if the 5 qubits of this QuantumChar are measured to
``5 = 00101``, the out come will be displayed as f.
"""
for i in range(2**self.size):
if self.decoder(i) == value:
return i
raise Exception("Value " + str(value) + " not supported by encoder.")
[docs] def encode(self, value, permit_dirtyness = False):
"""
The encode method allows to quickly bring a QuantumVariable in a desired
computational basis state.
A shorthand for this method is given by the ``[:]`` operator.
Note that the qubits to initialize have to be fresh (i.e. no operations
performed on them).
Parameters
----------
value :
A value supported by the encoder.
permit_dirtyness : bool, optional
Surpresses the error message when calling encode on dirty qubits.
Returns
-------
None.
Examples
--------
We create two quantum floats and encode the value 2.5. For one of them, we
perform an x gate onto the corresponding qubits, resulting in an error.
>>> from qrisp import QuantumFloat, x
>>> qf_0 = QuantumFloat(3, -1, signed = False)
>>> qf_1 = QuantumFloat(3, -1, signed = False)
>>> x(qf_0)
>>> qf_0.encode(2.5)
Exception: Tried to initialize qubits which are not fresh anymore.
>>> qf_1[:] = 2.5
>>> print(qf_1)
{2.5: 1.0}
"""
from qrisp.misc import check_if_fresh, int_encoder
if not permit_dirtyness:
if not check_if_fresh(self.reg, self.qs):
raise Exception("Tried to initialize qubits which are not fresh anymore.")
int_encoder(self, self.encoder(value))
[docs] def init_state(self, state_dic):
r"""
The ``init_state`` method allows the initialization of arbitrary quantum states.
It recieves a dictionary of the type
**{value : complex number}**
and initializes the **normalized** state. Amplitudes not specified are assumed
to be zero.
Note that the state initialization algorithm requires it's qubits to be in
state $\ket{0}$.
A shorthand for this method is the ``[:]`` operator, when handed the
corresponding dictionary
Parameters
----------
state_dic : dict
Dictionary describing the wave function to be initialized.
Raises
------
Exception
Tried to initialize qubits which are not fresh anymore.
Examples
--------
We create a QuantumFloat and encode the state
.. math::
\ket{\psi} = \sqrt{\frac{1}{3}} \ket{0.5} + i\sqrt{\frac{2}{3}} \ket{2}
>>> from qrisp import QuantumFloat
>>> qf = QuantumFloat(3, -1)
We can now use either
>>> qf.init_state({0.5: (1/3)**0.5, 2.0 : 1j*(2/3)**0.5})
or:
>>> qf[:] = {0.5: (1/3)**0.5, 2.0 : 1j*(2/3)**0.5}
To acquire the expected result
>>> print(qf)
{2.0: 0.6667, 0.5: 0.3333}
"""
from qrisp.misc import check_if_fresh
if not check_if_fresh(self.reg, self.qs):
raise Exception("Tried to initialize qubits which are not fresh anymore.")
from qrisp import init_state
target_array = np.zeros(2**self.size, dtype=np.complex128)
for key in state_dic.keys():
target_array[self.encoder(key)] = state_dic[key]
target_array = target_array / np.vdot(target_array, target_array) ** 0.5
init_state(self, target_array)
def append(self, operation):
self.qs.append(operation, self)
[docs] def extend(self, amount, position=-1):
"""
This method is used to add more qubits to the QuantumVariable. Using the
position keyword it is possible to specify the position where the qubits should
be added. By default, the qubits are added at the end.
Parameters
----------
amount : int
The amount of qubits to add.
position : int, optional
The position of where to add the qubits. By default, qubits are added at the
end.
st of qubits which are to be added to the QuantumVariable.
The default is None.
Raises
------
Exception
Missmatch between proposed qubits and amount integer.
Returns
-------
None.
Examples
--------
We create a QuantumVariable and extend it with some extra qubits.
>>> from qrisp import QuantumVariable
>>> qv = QuantumVariable(3)
>>> print(qv.reg)
[Qubit(qv.0), Qubit(qv.1), Qubit(qv.2)]
>>> qv.extend(3)
>>> print(qv.reg)
[Qubit(qv.0), Qubit(qv.1), Qubit(qv.2), Qubit(qv.6), Qubit(qv.6), Qubit(qv.6)]
"""
if position == -1:
position = self.size
insertion_qubits = self.qs.request_qubits(amount)
for i in range(amount):
insertion_qubits[i].identifier = self.name + "_ext_" + str(self.qs.qubit_index_counter[0]) + "." + str(self.size)
self.reg.insert(position + i, insertion_qubits[i])
self.size += 1
[docs] def reduce(self, qubits, verify=False):
r"""
Reduces the qubit count of the QuantumVariable by removing a specified set of
qubits.
Parameters
----------
qubits : list
The qubits to remove from the QuantumVariable.
verify : bool
Boolean value which indicates wether Qrisp should verify that the reduced
qubits are in the $\ket{0}$ state.
Raises
------
Exception
Qubits not present in QuantumVariable.
Exception
Verification that the given qubits are in $\ket{0}$ state failed.
Examples
--------
We create a QuantumVariable with 5 qubits and remove the first 2
>>> from qrisp import QuantumVariable
>>> qv = QuantumVariable(5)
>>> print(qv.reg)
[Qubit(qv.0), Qubit(qv.1), Qubit(qv.2), Qubit(qv.3), Qubit(qv.4)]
>>> qv.reduce(qv[:2])
>>> print(qv.reg)
[Qubit(qv.2), Qubit(qv.3), Qubit(qv.4)]
"""
try:
len(qubits)
except TypeError:
qubits = [qubits]
if not set(qubits).issubset(self.reg):
raise Exception("Tried to reduce QuantumVariable by invalid qubits")
# Find Qubits to be cleared
for i in range(len(qubits)):
for j in range(self.size):
if self.reg[j] == qubits[i]:
self.reg[j].identifier = "reduced_" + str(self.qs.qubit_index_counter[0])
self.qs.qubit_index_counter += 1
self.reg.pop(j)
break
self.qs.clear_qubits(qubits, verify)
# Adjust variable size
self.size -= len(qubits)
[docs] def get_measurement(
self,
plot=False,
backend=None,
shots=10000,
compile=True,
compilation_kwargs={},
subs_dic={},
circuit_preprocessor=None,
filename=None,
precompiled_qc = None
):
r"""
Method for quick access to the measurement results of the state of the variable.
This method returns a dictionary of the type {value : p} where p indicates the
probability with which that value is measured.
Parameters
----------
plot : Bool, optional
Plots the measurement results as a historgram. The default is False.
backend : BackendClient, optional
The backend on which to evaluate the quantum circuit. The default can be
specified in the file default_backend.py.
shots : integer, optional
The amount of shots to evaluate the circuit. The default is 10000.
compile : bool, optional
Boolean indicating if the .compile method of the underlying QuantumSession
should be called before. The default is True.
compilation_kwargs : dict, optional
Keyword arguments for the compile method. For more details check
:meth:`QuantumSession.compile <qrisp.QuantumSession.compile>`. The default
is ``{}``.
subs_dic : dict, optional
A dictionary of Sympy symbols and floats to specify parameters in the case
of a circuit with unspecified, :ref:`abstract parameters<QuantumCircuit>`.
The default is {}.
circuit_preprocessor : Python function, optional
A function which recieves a QuantumCircuit and returns one, which is applied
after compilation and parameter substitution. The default is None.
filename : string, optional
The location of where to save a generated plot. The default is None.
Raises
------
Exception
If the containing QuantumSession is in a quantum environment, it is not
possible to execute measurements.
Returns
-------
dict
A dictionary of values and their corresponding measurement probabilities.
Examples
--------
We create an integer :ref:`QuantumFloat`, encode the value 1 and bring the qubit
with significance 2 in superposition. We utilize the Qiskit transpiler by
transpiling into the gate set $\{\text{CX}, \text{U}\}$
>>> from qrisp import QuantumFloat, h
>>> qf = QuantumFloat(3,-1)
>>> qf[:] = 1
>>> h(qf[2])
>>> mes_results = qf.get_measurement(transpilation_kwargs = {"basis_gates" : ["cx", "u"]}) # noqa:501
>>> print(mes_results)
{1.0: 0.5, 3.0: 0.5}
"""
if backend is None:
if self.qs.backend is None:
from qrisp.default_backend import def_backend
backend = def_backend
else:
backend = self.qs.backend
if len(self.qs.env_stack) != 0:
raise Exception("Tried to get measurement within open environment")
if self.is_deleted():
raise Exception("Tried to get measurement from deleted QuantumVariable")
if self.size == 0:
return {"": 1.0}
if precompiled_qc is None:
if compile:
qc = qompiler(
self.qs, intended_measurements=self.reg, **compilation_kwargs
)
else:
qc = self.qs.copy()
else:
qc = precompiled_qc.copy()
# Bind parameters
if subs_dic:
qc = qc.bind_parameters(subs_dic)
from qrisp.core.compilation import combine_single_qubit_gates
qc = combine_single_qubit_gates(qc)
# Copy circuit in over to prevent modification
# from qrisp.quantum_network import QuantumNetworkClient
# if isinstance(backend, QuantumNetworkClient):
# self.qs.data = []
# shots = 1
# Execute user specified circuit_preprocessor
if circuit_preprocessor is not None:
qc = circuit_preprocessor(qc)
qc = qc.transpile()
from qrisp.misc import get_measurement_from_qc
counts = get_measurement_from_qc(qc, self.reg, backend, shots)
# Insert outcome labels (if available and hashable)
try:
new_counts_dic = {}
sorted_keys = list(counts.keys())
sorted_keys.sort()
for key in sorted_keys:
new_counts_dic[self.decoder(key)] = counts[key]
counts = new_counts_dic
# Sort keys
sorted_key_list = list(counts.keys())
sorted_key_list.sort(key=lambda x: -counts[x])
counts = {key: counts[key] for key in sorted_key_list}
except TypeError:
counts_tuple_list = []
for key in counts.keys():
counts_tuple_list.append((key, counts[key]))
counts = counts_tuple_list
counts.sorted(key=lambda x: x[1])
if plot:
outcome_labels = []
for i in range(2**self.size):
temp = self.decoder(i)
try:
hash(temp)
except TypeError:
raise Exception(
"Outcome value " + str(self.decoder(i)) + " is not hashable"
)
outcome_labels.append(temp)
plot_histogram(outcome_labels, counts, filename)
plt.show()
# Return dictionary of measurement results
return counts
[docs] def most_likely(self, **kwargs):
"""
Performs a measurement and returns the most likely outcome.
Parameters
----------
**kwargs : Keyword arguments for the get_measurement call.
Examples
--------
>>> from qrisp import QuantumFloat, ry
>>> import numpy as np
>>> qf = QuantumFloat(3)
>>> ry(np.pi*9/8, qf[0])
>>> print(qf)
{1: 0.9619, 0: 0.0381}
>>> qf.most_likely()
1
"""
return list(self.get_measurement())[0]
def __getitem__(self, key):
return self.reg[key]
def __str__(self):
return str(self.get_measurement())
def __repr__(self):
return "<" + str(type(self)).split(".")[-1][:-2] + " '" + self.name + "'>"
return str(type(self)).split(".")[-1][:-2] + "(name = " + self.name + ")"
return str(self)
def __del__(self):
i = 0
while i < len(self.live_qvs):
if self.live_qvs[i]() is None or id(self) == id(self.live_qvs[i]()):
self.live_qvs.pop(i)
continue
i += 1
def __len__(self):
return self.size
# Overload equality operator to use python syntax for if environments?
# Not sure if the possible user confusion is worth it
def __eq__(self, other):
from qrisp.environments import q_eq
return q_eq(self, other)
def __ne__(self, other):
from qrisp.environments import q_eq
return q_eq(self, other, invert = True)
def __hash__(self):
return self.creation_time
def __setitem__(self, key, value):
if key != slice(None, None, None):
raise Exception(
"Tried to encode value into QuantumVariable using non-trivial slicing."
)
if isinstance(type(value), type(None)):
return
if isinstance(value, dict):
self.init_state(value)
return
if isinstance(value, QuantumVariable):
self.init_from(value)
return
self.encode(value)
[docs] def app_phase_function(self, phi):
r"""
Applies a previously specified phase function to each computational basis state
of the QuantumVariable using Gray-Synthesis.
For a given phase function $\phi(x)$ and a QuantumVariable in state
$\ket{\psi} = \sum_{x \in \text{Labels}} a_x \ket{x}$ this method acts as:
.. math::
U_{\phi} \sum_{x \in \text{Labels}} a_x \ket{x} =
\sum_{x \in \text{Labels}} \text{exp}(i\phi(x)) a_x \ket{x}
Parameters
----------
phi : Python function
A Python function which turns the labels of the QuantumVariable into floats.
Examples
--------
We create a QuantumFloat and encode the k-th basis state of the Fourier basis.
Finally, we will apply an inverse Fourier transformation to measure k in the
computational basis.
>>> import numpy as np
>>> from qrisp import QuantumFloat, h, QFT
>>> n = 5
>>> qf = QuantumFloat(n, signed = False)
>>> h(qf)
After this, qf is in the state
.. math::
\ket{\text{qf}} = \frac{1}{\sqrt{2^n}} \sum_{x = 0}^{2^n} \ket{x}
We specify phi
>>> k = 4
>>> def phi(x):
>>> return 2*np.pi*x*k/2**n
And apply phi as a phase function
>>> qf.app_phase_function(phi)
qf is now in the state
.. math::
\ket{\text{qf}} = \frac{1}{\sqrt{2^n}} \sum_{x = 0}^{2^n}
\text{exp}\left( \frac{2\pi ikx}{2^n}\right) \ket{x}
Finally we apply the inverse Fourier transformation and measure:
>>> QFT(qf, inv = True)
>>> print(qf)
{4: 1.0}
"""
from qrisp.misc import app_phase_function
app_phase_function([self], phi)
[docs] def uncompute(self, do_it=True, recompute=False):
"""
Method for automatic uncomputation. Uses a generalized form of
`this algorithm <https://dl.acm.org/doi/10.1145/3453483.3454040>`_.
For more information check the
:ref:`uncomputation documentation<uncomputation>`.
Parameters
----------
do_it : bool, optional
If set to False, this variable will be appended to the uncomputation stack
of it's QuantumSession and uncomputed once an uncompute call with
``do_it = True`` is performed. The default is True.
recompute : bool, optional
If set to True, this QuantumVariable will be uncomputed but temporarily
recomputed, if it is required for the uncomputation of another
QuantumVariable. For more information check
:ref:`recomputation <recomputation>`. The default is False.
Examples
--------
We create two QuantumVariables, apply some gates and perform automatic
uncomputation:
>>> from qrisp import QuantumVariable, x, cx, h, p, mcx
>>> a = QuantumVariable(3)
>>> b = QuantumVariable(2)
>>> mcx(a, b[0])
>>> h(a[:2])
>>> x(b[0])
>>> cx(b[0], b[1])
>>> p(0.5, b[1])
>>> print(a.qs)
::
QuantumCircuit:
--------------
┌───┐
a.0: ──■──┤ H ├───────────────
│ ├───┤
a.1: ──■──┤ H ├───────────────
│ └───┘
a.2: ──■──────────────────────
┌─┴─┐┌───┐
b.0: ┤ X ├┤ X ├──■────────────
└───┘└───┘┌─┴─┐┌────────┐
b.1: ──────────┤ X ├┤ P(0.5) ├
└───┘└────────┘
Live QuantumVariables:
---------------------
QuantumVariable a
QuantumVariable b
>>> b.uncompute()
>>> print(b.qs)
::
QuantumCircuit:
--------------
┌────────┐ ┌────────┐┌───┐
a.0: ┤0 ├──────────────────────────────┤0 ├┤ H ├
│ │ │ │├───┤
a.1: ┤1 ├──────────────────────────────┤1 ├┤ H ├
│ pt3cx │ │ pt3cx │└───┘
a.2: ┤2 ├──────────────────────────────┤2 ├─────
│ │┌───┐ ┌───┐│ │
b.0: ┤3 ├┤ X ├──■──────────────■──┤ X ├┤3 ├─────
└────────┘└───┘┌─┴─┐┌────────┐┌─┴─┐└───┘└────────┘
b.1: ───────────────┤ X ├┤ P(0.5) ├┤ X ├────────────────────
└───┘└────────┘└───┘
Live QuantumVariables:
---------------------
QuantumVariable a
"""
if self.is_deleted():
raise Exception("Tried to uncompute deleted QuantumVariable")
if do_it:
from qrisp.uncomputation import uncompute
uncompute(self.qs, self.qs.uncomp_stack + [self], recompute)
self.qs.uncomp_stack = []
else:
self.qs.uncomp_stack.append(self)
def get_unique_name(self, name=None):
if name is None:
from qrisp import QuantumBool, QuantumChar, QuantumFloat
if isinstance(self, QuantumBool):
name = "qbl"
elif isinstance(self, QuantumFloat):
name = "qf"
elif isinstance(self, QuantumChar):
name = "qch"
else:
name = "qv"
while True:
try:
naming_number = self.name_tracker[name]
self.name_tracker[name] += 1
name = name + "_" + str(naming_number)
except KeyError:
self.name_tracker[name] = 1
name = name + "_0"
i = 0
while i < len(QuantumVariable.live_qvs):
qv = QuantumVariable.live_qvs[i]()
if qv is None:
QuantumVariable.live_qvs.pop(i)
continue
if qv.name == name:
break
i += 1
else:
break
return name
[docs] def init_from(self, other):
r"""
Method to initiate a QuantumVariable based on the state of another. This method
does NOT copy the state. Much rather it performs the operation
.. math::
U_{\text{init_from}} \left( \sum_{x \in \text{labels}} a_x \ket{x}
\right) \ket{0} = \sum_{x \in \text{labels}} a_x \ket{x} \ket{x}
This is different from a state copying operation:
.. math::
U_{\text{copy}} \left( \sum_{x \in \text{labels}} a_x \ket{x} \right)
\ket{0} = \left( \sum_{x \in \text{labels}} a_x \ket{x} \right)
\left( \sum_{x \in \text{labels}} a_x \ket{x} \right)
A shorthand for initiating this way is the ``[:]`` operator.
Parameters
----------
other : QuantumVariable
The QuantumVariable from which to initiate.
Raises
------
Exception
Tried to initialize qubits which are not fresh anymore.
Examples
--------
We create a QuantumFloat, and bring it into superposition.
>>> from qrisp import QuantumFloat, h, multi_measurement
>>> qf_a = QuantumFloat(8)
>>> qf_a[:] = 6
>>> h(qf_a[0])
>>> print(qf_a)
{6: 0.5, 7: 0.5}
We now duplicate and initiate the duplicate
>>> qf_b = qf_a.duplicate()
>>> print(qf_b)
{0: 1.0}
>>> qf_b.init_from(qf_a)
>>> print(multi_measurement([qf_a, qf_b]))
{(6, 6): 0.5, (7, 7): 0.5}
The slicing operator achieves the same:
>>> qf_c = qf_a.duplicate()
>>> qf_c[:] = qf_b
>>> print(multi_measurement([qf_a, qf_b, qf_c]))
{(6, 6, 6): 0.5, (7, 7, 7): 0.5}
"""
if not type(self) == type(other):
raise Exception(
"Tried to initialize " + str(type(self)) + " from " + str(type(other))
)
from qrisp.misc import check_if_fresh
if not check_if_fresh(self.reg, self.qs):
raise Exception("Tried to initialize qubits which are not fresh anymore")
self.qs.cx(other.reg, self.reg)
[docs] @classmethod
def custom(self, label_list, decoder=None, qs=None, name=None):
"""
Creates a QuantumVariable with customized outcome labels.
Note that this is a class method, implying there is no need to create another
QuantumVariable first to call this method.
Parameters
----------
label_list : list
A list of outcome labels.
decoder : function, optional
The decoder function. If given none, the labels will be encoded according to
their placement in the ``label_list``.
qs : QuantumSession, optional
The :ref:`QuantumSession` in which to register the customized
QuantumVariable. If given none, the QuantumVariable will be registered in a
new QuantumSession.
name : string, optional
The name of the QuantumVariable. If given none, a suited name will be
generated.
Returns
-------
CustomQuantumVariable
A QuantumVariable with the desired outcome labels.
Examples
--------
We create a QuantumVariable with some examples values as outcome labels and
bring it into uniform superposition.
>>> from qrisp import QuantumVariable, h
>>> qv = QuantumVariable.custom(["lorem", "ipsum", "dolor", "sit", 42, (1,2,3)])
>>> h(qv)
>>> print(qv)
{'lorem': 0.125, 'ipsum': 0.125, 'dolor': 0.125, 'sit': 0.125, 42: 0.125,
(1, 2, 3): 0.125, 'undefined_label_6': 0.125, 'undefined_label_7': 0.125}
"""
from qrisp.misc import custom_qv
return custom_qv(label_list, decoder=decoder, qs=qs, name=name)
def plot_histogram(outcome_labels, counts, filename=None):
res_list = []
for k in range(len(outcome_labels)):
try:
res_list.append(counts[outcome_labels[k]])
except KeyError:
res_list.append(0)
plt.bar(outcome_labels, res_list, width = 0.8/len(outcome_labels))
plt.grid()
plt.ylabel("Measurement probability")
plt.xlabel("QuantumVariable value")
if filename:
plt.savefig(filename, dpi=400, bbox_inches="tight")
else:
plt.show()