Closed adrianor closed 10 months ago
Thank you @adrianor for reporting this.
The problem here arises from a minor difference in the signature of the controlled unitary gate in Qiskit and in qiskit-symb
. In Qiskit, the CUGate
constructor (and the corresponding QuantumCircuit.cu
method you are using in your code) takes 3 angles theta, phi, lam + the global phase gamma while, in qiskit-symb
, the last parameter gamma is omitted.
From my point of view, this is actually a small but relevant inconsistency in Qiskit, since CUGate
is the only gate that takes one extra "angle parameter" compared to its corresponding UGate
(not controlled) and indeed several open issues on the main qiskit-terra
repository are related to this problem (e.g. https://github.com/Qiskit/qiskit-terra/issues/9763, https://github.com/Qiskit/qiskit-terra/issues/7326).
However, this is for sure not handled correctly in qiskit-symb
and the error message is very misleading. So, waiting for the problem to be properly solved in Qiskit itself, I can temporarily fix qiskit-symb
so that it shows an appropriate error message by marking the CUGate
as "not implemented".
Meanwhile, you can use the Qiskit transpiler to work around the problem and run your code with no errors:
import numpy as np
from qiskit import transpile
from qiskit.circuit import QuantumRegister, ClassicalRegister, QuantumCircuit, Parameter
from qiskit_symb.utils import get_symbolic_gates_names
st0 = Parameter("st0")
st1 = Parameter("st1")
def u1red_cr_overrot(alpha, theta):
qreg_q = QuantumRegister(2, "q")
creg_c = ClassicalRegister(2, "c")
circuit = QuantumCircuit(qreg_q, creg_c)
circuit.u(1.570796326794897, -2.9755113931440857, 4.71238898038469 + st1, qreg_q[1])
circuit.u(1.8942714348583476, -4.71238898038469, 0.0 + st0, qreg_q[0])
circuit.cu(np.pi / 2, np.pi / 2, -np.pi / 2, 0, qreg_q[1], qreg_q[0])
symbolic_gates = get_symbolic_gates_names()
symbolic_gates.remove("cu")
circuit = transpile(circuit, basis_gates=symbolic_gates)
return qreg_q, creg_c, circuit
In this case, the CUGate
transforms into a CU3Gate
but the final circuit is always exactly equivalent to the original. Note that this also works correctly when the global phase parameter gamma is not 0.
There is an issue when one tries to use the package and the circuit contains controlled unitaries. For instance the circuit
raises no problem when one draws it:
But if one tries to get its
Statevector
formit raises the following error message:
Note that this also happens if one wants to obtain its
Operator
form: