The problem of encountering negative eigenvalues in the density matrix during Quantum State Tomography (QST) is well-explained in this Quantum Computing Stack Exchange post.
However, I am seeking clarification on applying this concept to a 10-qubit code and performing QST on a 3-qubit subsystem of it. My primary objective is to compute the entropy, but the negative eigenvalues are causing divergence in my calculations.
When I execute QST using the following code in Cirq, I end up with a density matrix that has negative eigenvalues. Is there a theoretical method ensuring that QST results in a positive semi-definite density matrix?
cirq.experiments.state_tomography(
sampler: 'cirq.Sampler',
qubits: Sequence['cirq.Qid'],
circuit: 'cirq.Circuit',
repetitions: int = 1000,
prerotations: Optional[Sequence[Tuple[float, float]]] = None,
)
