Skip to main contentIBM Quantum Documentation
You are viewing the API reference for an old version of Qiskit SDK. Switch to latest version

ProcessTomographyFitter

ProcessTomographyFitter(result, circuits, meas_basis='Pauli', prep_basis='Pauli') GitHub(opens in a new tab)

Maximum-Likelihood estimation process tomography fitter.

Initialize tomography fitter with experimental data.

Parameters

  • result (Result) – a Qiskit Result object obtained from executing tomography circuits.
  • circuits (Union[List[QuantumCircuit], List[str]]) – a list of circuits or circuit names to extract count information from the result object.
  • meas_basis (Union[TomographyBasis, str]) – (default: ‘Pauli’) A function to return measurement operators corresponding to measurement outcomes. See Additional Information.
  • prep_basis (Union[TomographyBasis, str]) – (default: ‘Pauli’) A function to return preparation operators. See Additional Information

Attributes

data

Return tomography data

measure_basis

Return the tomography measurement basis.

preparation_basis

Return the tomography preparation basis.


Methods

add_data

ProcessTomographyFitter.add_data(result, circuits)

Add tomography data from a Qiskit Result object.

Parameters

  • result (Result) – a Qiskit Result object obtained from executing tomography circuits.
  • circuits (list) – a list of circuits or circuit names to extract count information from the result object.

fit

ProcessTomographyFitter.fit(method='auto', standard_weights=True, beta=0.5, **kwargs)

Reconstruct a quantum channel using CVXPY convex optimization.

Choi matrix

The Choi matrix object is a QuantumChannel representation which may be converted to other representations using the classes SuperOp, Kraus, Stinespring, PTM, Chi from the module qiskit.quantum_info.operators. The raw matrix data for the representation may be obtained by channel.data.

Fitter method

The cvx fitter method used CVXPY convex optimization package. The lstsq method uses least-squares fitting (linear inversion). The auto method will use cvx if the CVXPY package is found on the system, otherwise it will default to lstsq.

Objective function

This fitter solves the constrained least-squares minimization: minimize:axb2minimize: ||a \cdot x - b ||_2

subject to:

  • x>>0x >> 0 (PSD)
  • trace(x)=dim\text{trace}(x) = \text{dim} (trace)
  • partial_trace(x)=identity\text{partial\_trace}(x) = \text{identity} (trace_preserving)

where:

  • a is the matrix of measurement operators a[i]=vec(Mi).Ha[i] = \text{vec}(M_i).H
  • b is the vector of expectation value data for each projector b[i]Tr[Mi.Hx]=(ax)[i]b[i] \sim \text{Tr}[M_i.H \cdot x] = (a \cdot x)[i]
  • x is the vectorized Choi-matrix to be fitted

PSD constraint

The PSD keyword constrains the fitted matrix to be postive-semidefinite. For the lstsq fitter method the fitted matrix is rescaled using the method proposed in Reference [1]. For the cvx fitter method the convex constraint makes the optimization problem a SDP. If PSD=False the fitted matrix will still be constrained to be Hermitian, but not PSD. In this case the optimization problem becomes a SOCP.

Trace constraint

The trace keyword constrains the trace of the fitted matrix. If trace=None there will be no trace constraint on the fitted matrix. This constraint should not be used for process tomography and the trace preserving constraint should be used instead.

Trace preserving (TP) constraint

The trace_preserving keyword constrains the fitted matrix to be TP. This should only be used for process tomography, not state tomography. Note that the TP constraint implicitly enforces the trace of the fitted matrix to be equal to the square-root of the matrix dimension. If a trace constraint is also specified that differs from this value the fit will likely fail. Note that this can only be used for the CVX method.

CVXPY Solvers:

Various solvers can be called in CVXPY using the solver keyword argument. See the CVXPY documentation(opens in a new tab) for more information on solvers.

References:

[1] J Smolin, JM Gambetta, G Smith, Phys. Rev. Lett. 108, 070502

(2012). Open access: arXiv:1106.5458 [quant-ph].

Parameters

  • method (str) – (default: ‘auto’) the fitter method ‘auto’, ‘cvx’ or ‘lstsq’.
  • standard_weights (bool) – (default: True) apply weights to tomography data based on count probability
  • beta (float) – (default: 0.5) hedging parameter for converting counts to probabilities
  • **kwargs – kwargs for fitter method.

Raises

  • ValueError – In case the input data is no a valid process matrix
  • QiskitError – If the fit method is unrecognized

Returns

The fitted Choi-matrix J for the channel that maximizes basis_matrixvec(J)data2||\text{basis\_matrix} \cdot \text{vec}(J) - \text{data}||_2. The Numpy matrix can be obtained from Choi.data.

Return type

Choi

set_measure_basis

ProcessTomographyFitter.set_measure_basis(basis)

Set the measurement basis

Parameters

basis (Union[TomographyBasis, str]) – measurement basis

Raises

QiskitError – In case of invalid measurement or preparation basis.

set_preparation_basis

ProcessTomographyFitter.set_preparation_basis(basis)

Set the preparation basis function

Parameters

basis (Union[TomographyBasis, str]) – preparation basis

Raises

QiskitError – in case the basis has no preperation data

Was this page helpful?