pyHarm.Predictors.PredictorTangent

Classes

PredictorTangent

Define the tangent type of predictor. Using the Jacobian at solution point, a tangent to R(x)=0 solution is drawn and used as a prediction direction.

Module Contents

class pyHarm.Predictors.PredictorTangent.PredictorTangent(sign_ds, logger: logging.Logger | None = None, **kwargs)

Bases: pyHarm.Predictors.ABCPredictor.ABCPredictor

Define the tangent type of predictor. Using the Jacobian at solution point, a tangent to R(x)=0 solution is drawn and used as a prediction direction.

The tangent is computed using a QR decomposition of the Jacobian at the solution point.

predictor_name = 'Tangent Predictor'
factory_keyword: str = 'tangent'

keyword that is used to call the creation of this class in the system factory.

Type:

str

predict(sollist: list[pyHarm.Solver.SystemSolution], ds: float, k_imposed=None) → tuple[numpy.ndarray, pyHarm.Solver.SystemSolution, float]

Predicts the next starting point using the tangent.

Parameters:

  • sollist (list[SystemSolution]) – list of SystemSolution already solved during the analysis.

  • ds (float) – step size for the prediction.

  • k_imposed (None | int) – if not None, uses the k_imposed as the index of the last solution pointer.

Returns:

next predicted starting point. SystemSolution: last accepted point in the list of solutions. float: sign of the prediction used (-1 | 1)

Return type:

np.ndarray