Expected outcome of adding two PauliOp

Question

first of all, I would probably report this as a bug if I was confident enough in my QC abilities. As it stands, I've just started out, so I'm phrasing it as a question.

From what I can tell, adding two PauliOps in this specific case does not result in the expected operator matrix:

import qiskit as qk
from qiskit.opflow.primitive_ops import PauliOp

(PauliOp(qk.quantum_info.Pauli("X")) + PauliOp(qk.quantum_info.Pauli("iY"))).to_matrix()

results in

array([[0.+0.j, 1.-1.j],
       [1.+1.j, 0.+0.j]])

whereas the result that I would expect is the one i get when I run the following code:

PauliOp(qk.quantum_info.Pauli("X")).to_matrix() + PauliOp(qk.quantum_info.Pauli("iY")).to_matrix()

Result:

array([[0.+0.j, 2.+0.j],
       [0.+0.j, 0.+0.j]])

What further confuses me is that the deprecated version of PauliOp actually results in the expected operator. Running this code:

import qiskit as qk
from qiskit.aqua.operators.primitive_ops import PauliOp

(PauliOp(qk.quantum_info.Pauli("X")) + PauliOp(qk.quantum_info.Pauli("iY"))).to_matrix()

gives a deprecation warning and the [[0,2],[0,0]]-array that I want.

Has the behaviour of the addition changed or is this a bug? In any case, how can I construct the operator that I want (in a non-deprecated way)?

I'm running qiskit version 0.29.0

Thanks in advance!

Answer 1

You can do this instead:

(PauliOp(qk.quantum_info.Pauli("X")) + 1j*PauliOp(qk.quantum_info.Pauli("Y"))).to_matrix()

which gives what you expect:

array([[0.+0.j, 2.+0.j],
       [0.+0.j, 0.+0.j]])