Operator Evolutions (qiskit.opflow.evolutions
)¶
Evolutions are converters which traverse an Operator tree, replacing any EvolvedOp
e
with a Schrodinger equationstyle evolution CircuitOp
equalling or approximating the matrix exponential of i * the Operator contained inside
(e.primitive). The Evolutions are essentially implementations of Hamiltonian Simulation
algorithms, including various methods for Trotterization.
The EvolvedOp
is simply a placeholder signifying that the Operator inside it should be
converted to its exponential by the Evolution converter. All Operators
(not state_fns
) have
.exp_i()
methods which either return the exponential of the Operator directly,
or an EvolvedOp
containing the Operator.
Nota
Evolutions work with parameterized Operator coefficients, so
my_expectation.convert((t * H).exp_i())
, where t is a scalar or Terra Parameter and H
is an Operator, will produce a CircuitOp
equivalent to e^iHt.
Evolution Base Class¶
The EvolutionBase class gives an interface for algorithms to ask for Evolutions as
execution settings. For example, if an algorithm contains an Operator evolution step within it,
such as QAOA
, the algorithm can give the opportunity for the user
to pass an EvolutionBase of their choice to be used in that evolution step.
A base for Evolution converters. 
Evolutions¶
A factory class for convenient automatic selection of an Evolution algorithm based on the Operator to be converted. 

Class for wrapping Operator Evolutions for compilation ( 

Performs Evolution by classical matrix exponentiation, constructing a circuit with 

An Evolution algorithm replacing exponentiated sums of Paulis by changing them each to the Z basis, rotating with an rZ, changing back, and Trotterizing. 
Trotterizations¶
A base for Trotterization methods, algorithms for approximating exponentiations of operator sums by compositions of exponentiations. 

A factory for conveniently creating TrotterizationBase instances. 

Simple Trotter expansion, composing the evolution circuits of each Operator in the sum together 

Suzuki Trotter expansion, composing the evolution circuits of each Operator in the sum together by a recursive “bookends” strategy, repeating the whole composed circuit 

The QDrift Trotterization method, which selects each each term in the Trotterization randomly, with a probability proportional to its weight. 