Support parameterized operators

[mrossinek]

What should we add?

We would like to support parameterized operators. This would:

• use qiskit.circuit.Parameter to enable parameterization
• need to be supported in:
  • SparseLabelOp
  • PolynomialTensor (Parameter can actually be injected right into the matrices) (see Support ParameterExpression objects inside PolynomialTensor #894)
  • classes which contain those operators (see Support ParameterExpression objects inside PolynomialTensor #894)
• I think, this should not require any special handling in the QubitMapper and QubitConverter
• some mechanics may need to be added during our algorithms to deal with parameters (can be dealt of as a follow-up feature to this one)

This depends on:

• Parameter handling in SparsePauliOp Qiskit/qiskit#8620
• Hamiltonian classes being able to have their returned objects of second_q_op() be modifiable (i.e. parameterizable). I have some ideas for this but will provide more details once Implement FermionicOp.from_polynomial_tensor #666 has a PR; during its development I will pay particular attention to this feature. (see Support ParameterExpression objects inside PolynomialTensor #894)

[kevinsung]

You can assign this to me.

[kevinsung]

@mrossinek FYI I currently do not plan to work on adding Parameter support for PolynomialTensor, I suggest renaming this issue so it's specific to SparseLabelOp and creating a separate one for PolynomialTensor if necessary.