tequila-basic 1.9.8__py3-none-any.whl → 1.9.10__py3-none-any.whl

tequila/objective/objective.py

@@ -1,4 +1,7 @@
-import typing, copy, numbers
+import typing
+import copy
+import numbers
+from typing import Union
 from tequila.grouping.compile_groups import compile_commuting_parts
 from tequila import TequilaException
 from tequila.utils import JoinedTransformation
@@ -10,6 +13,7 @@ import collections
 
 # convenience
 
+
 class ExpectationValueImpl:
     """
     Implements the (uncompiled) Expectation Value as a class. Should not be called directly.
@@ -113,9 +117,12 @@ class ExpectationValueImpl:
         the ExpectationValueImpl structure with mapped qubits
 
         """
-        return ExpectationValueImpl(H=tuple([H.map_qubits(qubit_map=qubit_map) for H in self.H]),
-                                    U=self.U.map_qubits(qubit_map=qubit_map), contraction=self._contraction,
-                                    shape=self._shape)
+        return ExpectationValueImpl(
+            H=tuple([H.map_qubits(qubit_map=qubit_map) for H in self.H]),
+            U=self.U.map_qubits(qubit_map=qubit_map),
+            contraction=self._contraction,
+            shape=self._shape,
+        )
 
     def map_variables(self, variables: dict, *args, **kwargs):
         """
@@ -129,17 +136,25 @@ class ExpectationValueImpl:
         Circuit with changed variables
 
         """
-        return ExpectationValueImpl(H=self.H, U=self.U.map_variables(variables=variables, *args, **kwargs),
-                                    contraction=self._contraction, shape=self._shape)
+        return ExpectationValueImpl(
+            H=self.H,
+            U=self.U.map_variables(variables=variables, *args, **kwargs),
+            contraction=self._contraction,
+            shape=self._shape,
+        )
 
     def __call__(self, *args, **kwargs):
         raise TequilaException(
-            "Tried to call uncompiled ExpectationValueImpl, compile your objective before calling with tq.compile(objective) or evaluate with tq.simulate(objective)")
+            "Tried to call uncompiled ExpectationValueImpl, compile your objective before calling with tq.compile(objective) or evaluate with tq.simulate(objective)"
+        )
 
     def info(self, short=True, *args, **kwargs):
         if short:
-            print("Expectation Value with {qubits} active qubits and {paulis} paulistrings".format(
-                qubits=len(self.U.qubits), paulis=len(self.H)))
+            print(
+                "Expectation Value with {qubits} active qubits and {paulis} paulistrings".format(
+                    qubits=len(self.U.qubits), paulis=len(self.H)
+                )
+            )
         else:
             print("Hamiltonian:\n", str(self.H))
             print("\n", str(self.U))
@@ -256,7 +271,7 @@ class Objective:
         """
         variables = []
         for arg in self.args:
-            if hasattr(arg, 'extract_variables'):
+            if hasattr(arg, "extract_variables"):
                 variables += arg.extract_variables()
             else:
                 variables += []
@@ -303,7 +318,6 @@ class Objective:
 
     @property
     def args(self) -> typing.Tuple:
-
         if self._args is None:
             return tuple()
         else:
@@ -396,7 +410,7 @@ class Objective:
 
     def __invert__(self):
         new = Objective(args=[self])
-        return new ** -1
+        return new**-1
 
     @classmethod
     def unary_operator(cls, left, op):
@@ -417,8 +431,7 @@ class Objective:
             Objective representing op applied to objective left.
 
         """
-        return Objective(args=left.args,
-                         transformation=lambda *args: op(left.transformation(*args)))
+        return Objective(args=left.args, transformation=lambda *args: op(left.transformation(*args)))
 
     @classmethod
     def binary_operator(cls, left, right, op):
@@ -450,18 +463,23 @@ class Objective:
                 return cls.unary_operator(left=left, op=lambda E: op(E, right))
             else:
                 raise TequilaException(
-                    'BinaryOperator method called on types ' + str(type(left)) + ',' + str(type(right)))
+                    "BinaryOperator method called on types " + str(type(left)) + "," + str(type(right))
+                )
         elif isinstance(left, numbers.Number):
             if isinstance(right, Objective):
                 return cls.unary_operator(left=right, op=lambda E: op(left, E))
             else:
                 raise TequilaException(
-                    'BinaryOperator method called on types ' + str(type(left)) + ',' + str(type(right)))
+                    "BinaryOperator method called on types " + str(type(left)) + "," + str(type(right))
+                )
         else:
             split_at = len(left.args)
-            return Objective(args=left.args + right.args,
-                             transformation=JoinedTransformation(left=left.transformation, right=right.transformation,
-                                                                 split=split_at, op=op))
+            return Objective(
+                args=left.args + right.args,
+                transformation=JoinedTransformation(
+                    left=left.transformation, right=right.transformation, split=split_at, op=op
+                ),
+            )
 
     def wrap(self, op):
         """
@@ -540,12 +558,14 @@ class Objective:
 
         unique = self.count_expectationvalues(unique=True)
         measurements = self.count_measurements()
-        return "Objective with {} unique expectation values\n" \
-               "total measurements = {}\n" \
-               "variables          = {}\n" \
-               "types              = {}".format(unique, measurements, variables, types)
-
-    def __call__(self, variables=None, *args, **kwargs):
+        return (
+            "Objective with {} unique expectation values\n"
+            "total measurements = {}\n"
+            "variables          = {}\n"
+            "types              = {}".format(unique, measurements, variables, types)
+        )
+
+    def __call__(self, variables=None, initial_state=0, *args, **kwargs):
         """
         Return the output of the calculation the objective represents.
 
@@ -553,6 +573,8 @@ class Objective:
         ----------
         variables: dict:
             dictionary instantiating all variables that may appear within the objective.
+        initial_state: int or QubitWaveFunction:
+            the initial state of the circuit
         args
         kwargs
 
@@ -565,27 +587,28 @@ class Objective:
         # failsafe
         check_variables = {k: k in variables for k in self.extract_variables()}
         if not all(list(check_variables.values())):
-            raise TequilaException("Objective did not receive all variables:\n"
-                                   "You gave\n"
-                                   " {}\n"
-                                   " but the objective depends on\n"
-                                   " {}\n"
-                                   " missing values for\n"
-                                   " {}".format(variables, self.extract_variables(),
-                                                [k for k, v in check_variables.items() if not v]))
+            raise TequilaException(
+                "Objective did not receive all variables:\n"
+                "You gave\n"
+                " {}\n"
+                " but the objective depends on\n"
+                " {}\n"
+                " missing values for\n"
+                " {}".format(variables, self.extract_variables(), [k for k, v in check_variables.items() if not v])
+            )
 
         # avoid multiple evaluations
         evaluated = {}
         ev_array = []
         for E in self.args:
             if E not in evaluated:  #
-                expval_result = E(variables=variables, *args, **kwargs)
+                expval_result = E(variables=variables, initial_state=initial_state, *args, **kwargs)
                 evaluated[E] = expval_result
             else:
                 expval_result = evaluated[E]
             try:
                 expval_result = float(expval_result)
-            except:
+            except Exception:
                 pass  # allow array evaluation (non-standard operation)
             ev_array.append(expval_result)
         result = onp.asarray(self.transformation(*ev_array), dtype=float)
@@ -595,7 +618,7 @@ class Objective:
             return float(result[0])
         else:
             return result
-    
+
     def contract(self):
         """
         Exists only to be convient in optimizers, which all contract over VectrObjectives.
@@ -608,19 +631,20 @@ class Objective:
 
     def __len__(self):
         return 1
-    
+
     def is_translated(self):
         """
         check if the objective was already translated to a quantum backend
         """
         types = [type(E) for E in self.get_expectationvalues()]
         types = list(set(types))
-        if len(types)==0 or (ExpectationValueImpl in types and len(types)==1):
+        if len(types) == 0 or (ExpectationValueImpl in types and len(types) == 1):
             return False
         else:
             return True
 
-def ExpectationValue(U, H, optimize_measurements = False, *args, **kwargs) -> Objective:
+
+def ExpectationValue(U, H, optimize_measurements=False, *args, **kwargs) -> Objective:
     """
     Initialize an Objective which is just a single expectationvalue
     """
@@ -758,8 +782,10 @@ class Variable:
             new = Objective(args=[self, other], transformation=op)
         else:
             raise TequilaException(
-                "unknown type in left_helper of objective arithmetics with operation {}: {}".format(type(op),
-                                                                                                    type(other)))
+                "unknown type in left_helper of objective arithmetics with operation {}: {}".format(
+                    type(op), type(other)
+                )
+            )
         return new
 
     def _right_helper(self, op, other):
@@ -779,8 +805,10 @@ class Variable:
             new = Objective(args=[other, self], transformation=op)
         else:
             raise TequilaException(
-                "unknown type in left_helper of objective arithmetics with operation {}: {}".format(type(op),
-                                                                                                    type(other)))
+                "unknown type in left_helper of objective arithmetics with operation {}: {}".format(
+                    type(op), type(other)
+                )
+            )
         return new
 
     def __mul__(self, other):
@@ -796,7 +824,7 @@ class Variable:
         return self._left_helper(numpy.true_divide, other)
 
     def __neg__(self):
-        return Objective(args=[self], transformation=lambda v: numpy.multiply(v, -1.))
+        return Objective(args=[self], transformation=lambda v: numpy.multiply(v, -1.0))
 
     def __pow__(self, other):
         return self._left_helper(numpy.power, other)
@@ -815,7 +843,7 @@ class Variable:
 
     def __invert__(self):
         new = Objective(args=[self])
-        return new ** -1.0
+        return new**-1.0
 
     def __len__(self):
         return 1
@@ -827,7 +855,7 @@ class Variable:
             return True
 
     def apply(self, other):
-        assert (callable(other))
+        assert callable(other)
         return Objective(args=[self], transformation=other)
 
     def wrap(self, other):
@@ -838,6 +866,7 @@ class Variable:
 
     def toJson(self):
         import json
+
         return json.dumps(self, default=lambda o: o.__dict__)
 
 
@@ -857,15 +886,16 @@ class FixedVariable(float):
         return self
 
     def apply(self, other):
-        assert (callable(other))
+        assert callable(other)
         return Objective(args=[self], transformation=other)
 
     def wrap(self, other):
         return self.apply(other)
-    
+
     def map_variables(self, *args, **kwargs):
         return self
 
+
 def format_variable_list(variables: typing.List[typing.Hashable]) -> typing.List[Variable]:
     """
     Convenience functions to assign tequila variables.
@@ -885,8 +915,9 @@ def format_variable_list(variables: typing.List[typing.Hashable]) -> typing.List
         return [assign_variable(k) for k in variables]
 
 
-def format_variable_dictionary(variables: typing.Dict[typing.Hashable, typing.Any]) -> typing.Dict[
-    Variable, typing.Any]:
+def format_variable_dictionary(
+    variables: typing.Dict[typing.Hashable, typing.Any],
+) -> typing.Dict[Variable, typing.Any]:
     """
     Convenience function to assign tequila variables.
     Parameters
@@ -905,8 +936,9 @@ def format_variable_dictionary(variables: typing.Dict[typing.Hashable, typing.An
         return Variables(variables)
 
 
-def assign_variable(variable: typing.Union[typing.Hashable, numbers.Real, Variable, FixedVariable]) -> typing.Union[
-    Variable, FixedVariable]:
+def assign_variable(
+    variable: typing.Union[typing.Hashable, numbers.Real, Variable, FixedVariable],
+) -> typing.Union[Variable, FixedVariable]:
     """
     Convenience function; maps various objects into Variable, FixedVariable, or Variables, for easy duck-typing.
 
@@ -917,7 +949,7 @@ def assign_variable(variable: typing.Union[typing.Hashable, numbers.Real, Variab
 
     Raises
     ------
-    
+
     TequilaVariableException
 
 
@@ -942,8 +974,11 @@ def assign_variable(variable: typing.Union[typing.Hashable, numbers.Real, Variab
         return Variable(name=variable)
     else:
         raise TequilaVariableException(
-            "Only hashable types can be assigned to Variables. You passed down " + str(variable) + " type=" + str(
-                type(variable)))
+            "Only hashable types can be assigned to Variables. You passed down "
+            + str(variable)
+            + " type="
+            + str(type(variable))
+        )
 
 
 class Variables(collections.abc.MutableMapping):
@@ -985,5 +1020,5 @@ class Variables(collections.abc.MutableMapping):
         return result
 
     def __repr__(self):
-        xdict = {k:v for k,v in self.items()}
+        xdict = {k: v for k, v in self.items()}
         return xdict.__repr__()

tequila/objective/qtensor.py

@@ -1,7 +1,9 @@
-import numpy, typing
+import numpy
+import typing
 from .objective import Objective, ExpectationValueImpl, format_variable_dictionary
 from tequila import TequilaException
 
+
 class QTensor(numpy.ndarray):
     # see here: https://numpy.org/devdocs/user/basics.subclassing.html
 
@@ -15,17 +17,17 @@ class QTensor(numpy.ndarray):
         # do all-zero initialization
         shape = self.shape
         if objective_list is None:
-            with numpy.nditer(self, flags =["refs_ok"], op_flags=['readwrite']) as it:
+            with numpy.nditer(self, flags=["refs_ok"], op_flags=["readwrite"]) as it:
                 for x in it:
                     x[...] = Objective()
         else:
-            j=0
-            with numpy.nditer(self, flags =["refs_ok"], op_flags=['readwrite']) as it:
+            j = 0
+            with numpy.nditer(self, flags=["refs_ok"], op_flags=["readwrite"]) as it:
                 for x in it:
                     x[...] = objective_list[j]
-                    j=j+1
+                    j = j + 1
 
-    def __call__(self,variables=None, *args, **kwargs):
+    def __call__(self, variables=None, *args, **kwargs):
         """
         Return the output of the calculation the objective represents.
 
@@ -45,13 +47,15 @@ class QTensor(numpy.ndarray):
         # failsafe
         check_variables = {k: k in variables for k in self.extract_variables()}
         if not all(list(check_variables.values())):
-            raise TequilaException("Objective did not receive all variables:\n"
-                                   "You gave\n"
-                                   " {}\n"
-                                   " but the objective depends on\n"
-                                   " {}\n"
-                                   " missing values for\n"
-                                   " {}".format(variables, self.extract_variables(), [k for k,v in check_variables.items() if not v]))
+            raise TequilaException(
+                "Objective did not receive all variables:\n"
+                "You gave\n"
+                " {}\n"
+                " but the objective depends on\n"
+                " {}\n"
+                " missing values for\n"
+                " {}".format(variables, self.extract_variables(), [k for k, v in check_variables.items() if not v])
+            )
 
         # avoid multiple evaluations
         evaluated = {}
@@ -60,7 +64,7 @@ class QTensor(numpy.ndarray):
         for obj in newtensor:
             a = obj(variables=variables, *args, **kwargs)
             ev_array.append(a)
-        ev_array = numpy.reshape(ev_array,self.shape)
+        ev_array = numpy.reshape(ev_array, self.shape)
         if ev_array.shape == ():
             return float(ev_array)
         elif len(ev_array) == 1:
@@ -73,11 +77,11 @@ class QTensor(numpy.ndarray):
         _fn = numpy.vectorize(_f)
         return _fn(self)
 
-    def extract_variables(self)->list:
+    def extract_variables(self) -> list:
         newtensor = self.flatten()
         unique = []
         for obj in newtensor:
-            if hasattr(obj, 'extract_variables'):
+            if hasattr(obj, "extract_variables"):
                 var_list = obj.extract_variables()
                 for j in var_list:
                     if j not in unique:
@@ -94,7 +98,7 @@ class QTensor(numpy.ndarray):
         newtensor = self.flatten()
         expvals = []
         for obj in newtensor:
-            if hasattr(obj, 'get_expectationvalues'):
+            if hasattr(obj, "get_expectationvalues"):
                 expvals += obj.get_expectationvalues()
         return expvals
 
@@ -129,7 +133,7 @@ class QTensor(numpy.ndarray):
             return len(self.get_expectationvalues())
 
     def __repr__(self):
-        _repmat = numpy.empty(self.shape,dtype = object)
+        _repmat = numpy.empty(self.shape, dtype=object)
         _repmat = _repmat.flatten()
         newtensor = self.flatten()
         for i in range(len(newtensor)):
@@ -144,7 +148,7 @@ class QTensor(numpy.ndarray):
         newtensor = self.flatten()
         types = []
         for obj in newtensor:
-            if hasattr(obj, 'get_expectationvalues'):
+            if hasattr(obj, "get_expectationvalues"):
                 _types = [type(E) for E in obj.get_expectationvalues()]
                 for tt in _types:
                     types.append(tt)
@@ -157,20 +161,21 @@ class QTensor(numpy.ndarray):
 
         unique = self.count_expectationvalues(unique=True)
         measurements = self.count_measurements()
-        return "QTensor of shape {} with {} unique expectation values\n" \
-              "total measurements = {}\n" \
-              "variables          = {}\n" \
-              "types              = {}".format(self.shape, unique, measurements, variables, types)
+        return (
+            "QTensor of shape {} with {} unique expectation values\n"
+            "total measurements = {}\n"
+            "variables          = {}\n"
+            "types              = {}".format(self.shape, unique, measurements, variables, types)
+        )
 
     def contract(self):
         newtensor = self.flatten()
-        out_array=[obj for obj in newtensor]
+        out_array = [obj for obj in newtensor]
         summed = out_array[0]
         for entry in out_array[1:]:
             summed += entry
         return summed
 
-
     class HelperObject:
         """
         This is a small helper object class for tequila objectives
@@ -178,17 +183,21 @@ class QTensor(numpy.ndarray):
         create if like this:
             ff = HelperObject(func=f) where f is the function you want to apply later (e.g. numpy.sin)
         use if like this with tequila objectives
-            f_on_objective = ff(objective) 
+            f_on_objective = ff(objective)
         """
+
         def __init__(self, func):
             self.func = func
+
         def __call__(self, objective):
             return objective.apply(self.func)
 
+
 # ------------------------------------------------------
 # backward compatibility with old VectorObjective class
 # ------------------------------------------------------
 
+
 def vectorize(objectives):
     """
     Combine several objectives in order, into one longer vector.
@@ -206,6 +215,7 @@ def vectorize(objectives):
     """
     return QTensor(objective_list=objectives, shape=(len(objectives),))
 
+
 def VectorObjective(argsets: typing.Iterable = None, transformations: typing.Iterable[callable] = None):
     if argsets is None:
         return QTensor()
@@ -219,4 +229,3 @@ def VectorObjective(argsets: typing.Iterable = None, transformations: typing.Ite
             objective_list.append(Objective(args=argsets[i], transformation=transformations[i]))
 
     return vectorize(objectives=objective_list)
-

tequila/optimizers/__init__.py

@@ -6,7 +6,9 @@ from tequila.optimizers.optimizer_gd import minimize as minimize_gd
 from tequila.simulators.simulator_api import simulate
 from dataclasses import dataclass
 
-import typing, numbers, numpy
+import typing
+import numbers
+import numpy
 
 
 @dataclass
@@ -16,27 +18,26 @@ class _Optimizers:
     methods: list = None
 
 
-SUPPORTED_OPTIMIZERS = ['scipy', 'gpyopt', 'gd']
+SUPPORTED_OPTIMIZERS = ["scipy", "gpyopt", "gd"]
 INSTALLED_OPTIMIZERS = {}
-INSTALLED_OPTIMIZERS['scipy'] = _Optimizers(cls=OptimizerSciPy,
-                                            minimize=minimize_scipy,
-                                            methods=OptimizerSciPy.available_methods())
-INSTALLED_OPTIMIZERS['gd'] = _Optimizers(cls=OptimizerGD,
-                                         minimize=minimize_gd,
-                                         methods=OptimizerGD.available_methods())
+INSTALLED_OPTIMIZERS["scipy"] = _Optimizers(
+    cls=OptimizerSciPy, minimize=minimize_scipy, methods=OptimizerSciPy.available_methods()
+)
+INSTALLED_OPTIMIZERS["gd"] = _Optimizers(cls=OptimizerGD, minimize=minimize_gd, methods=OptimizerGD.available_methods())
 
 has_gpyopt = False
 try:
     from tequila.optimizers.optimizer_gpyopt import OptimizerGPyOpt
     from tequila.optimizers.optimizer_gpyopt import minimize as minimize_gpyopt
 
-    INSTALLED_OPTIMIZERS['gpyopt'] = _Optimizers(cls=OptimizerGPyOpt,
-                                                 minimize=minimize_gpyopt,
-                                                 methods=OptimizerGPyOpt.available_methods())
+    INSTALLED_OPTIMIZERS["gpyopt"] = _Optimizers(
+        cls=OptimizerGPyOpt, minimize=minimize_gpyopt, methods=OptimizerGPyOpt.available_methods()
+    )
     has_gpyopt = True
 except ImportError:
     has_gpyopt = False
 
+
 def show_available_optimizers(module=None):
     """
     Returns
@@ -65,13 +66,15 @@ def show_available_optimizers(module=None):
         print("Installed optimizer modules: ", list(INSTALLED_OPTIMIZERS.keys()))
 
 
-def minimize(objective,
-             method: str = "bfgs",
-             variables: list = None,
-             initial_values: typing.Union[dict, numbers.Number, typing.Callable] = 0.0,
-             maxiter: int = None,
-             *args,
-             **kwargs):
+def minimize(
+    objective,
+    method: str = "bfgs",
+    variables: list = None,
+    initial_values: typing.Union[dict, numbers.Number, typing.Callable] = 0.0,
+    maxiter: int = None,
+    *args,
+    **kwargs,
+):
     """
 
     Parameters
@@ -115,10 +118,12 @@ def minimize(objective,
 
     """
 
-    ovtmp=objective.extract_variables()
-    fast_return=False
+    ovtmp = objective.extract_variables()
+    fast_return = False
     if ovtmp is None or len(ovtmp) == 0:
-        return OptimizerResults(energy=float(simulate(objective, *args, **kwargs)), variables={}, history=OptimizerHistory())
+        return OptimizerResults(
+            energy=float(simulate(objective, *args, **kwargs)), variables={}, history=OptimizerHistory()
+        )
 
     for k, v in INSTALLED_OPTIMIZERS.items():
         if method.lower() in v.methods or method.upper() in v.methods:
@@ -128,7 +133,10 @@ def minimize(objective,
                 variables=variables,
                 initial_values=initial_values,
                 maxiter=maxiter,
-                *args, **kwargs)
+                *args,
+                **kwargs,
+            )
 
     raise TequilaOptimizerException(
-        "Could not find optimization method {} in tequila optimizers. You might miss dependencies")
+        "Could not find optimization method {} in tequila optimizers. You might miss dependencies"
+    )

tequila/optimizers/_containers.py

@@ -2,12 +2,12 @@ import numpy
 from tequila.objective import format_variable_dictionary
 from tequila.tools.qng import evaluate_qng
 import sys
+
 """
 Define Containers for SciPy usage
 """
 
 
-
 class _EvalContainer:
     """
     Container Class to access scipy and keep the optimization history.
@@ -35,8 +35,9 @@ class _EvalContainer:
 
     """
 
-    def __init__(self, objective, param_keys, passive_angles=None, samples=None, save_history=True,
-                 print_level: int = 3):
+    def __init__(
+        self, objective, param_keys, passive_angles=None, samples=None, save_history=True, print_level: int = 3
+    ):
         self.objective = objective
         self.samples = samples
         self.param_keys = param_keys
@@ -135,11 +136,14 @@ class _QngContainer(_EvalContainer):
         evaluate the qng.
     """
 
-
     def __init__(self, combos, param_keys, passive_angles=None, samples=None, save_history=True):
-
-        super().__init__(objective=None, param_keys=param_keys, passive_angles=passive_angles,
-                         samples=samples, save_history=save_history)
+        super().__init__(
+            objective=None,
+            param_keys=param_keys,
+            passive_angles=passive_angles,
+            samples=samples,
+            save_history=save_history,
+        )
 
         self.combos = combos