UGParameterEstimator.optimizers package
Submodules
UGParameterEstimator.optimizers.bayesOptimizer module
- class UGParameterEstimator.optimizers.bayesOptimizer.BayesOptimizer(parametermanager: ParameterManager, epsilon=0.0001, minreduction=0.0001, max_iterations=20)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.gainedLevMarOptimizer module
- class UGParameterEstimator.optimizers.gainedLevMarOptimizer.GainedLevMarOptimizer(maxiterations=15, tau=0.01, presteps=5, epsilon=0.001, minreduction=0.0001, initial_lam=None, differencing=Differencing.forward)
Bases:
Optimizer- calculateDelta(V, r, p, lam)
- calculateGainRatio(S, newS, delta, lam, grad)
- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.gaussNewtonOptimizer module
- class UGParameterEstimator.optimizers.gaussNewtonOptimizer.GaussNewtonOptimizer(linesearchmethod: LineSearch, maxiterations=15, epsilon=0.001, minreduction=0.0001, differencing=Differencing.forward)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.gradDescOptimizer module
- class UGParameterEstimator.optimizers.gradDescOptimizer.GradientDescentOptimizer(linesearchmethod: LineSearch, maxiterations=15, epsilon=0.001, minreduction=0.0001, max_error_ratio=(0.05, 0.95), differencing=Differencing.forward)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.levMarOptimizer module
- class UGParameterEstimator.optimizers.levMarOptimizer.LevMarOptimizer(maxiterations=15, initial_lam=0.01, nu=10, P=10, P_iteration_count=3, scaling=False, epsilon=0.001, minreduction=0.0001, differencing=Differencing.forward)
Bases:
Optimizer- calculateDelta(V, r, p, lam)
- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.mmaOptimizer module
- class UGParameterEstimator.optimizers.mmaOptimizer.MMAOptimizer(maximum, minimum, minreduction=0.0001, max_iterations=15, epsilon=0.001, differencing=Differencing.forward)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.optimizer module
- class UGParameterEstimator.optimizers.optimizer.Optimizer(epsilon, differencing: Differencing)
Bases:
ABCA base class for all optimizers, defining the interface and common helper methods
- class Differencing(value)
Bases:
EnumAn enumeration.
- central = 1
- forward = 2
- pure_central = 4
- pure_forward = 3
- getJacobiMatrix(point, evaluator, target, result)
Calculates the jacobi matrix in parallel using finite differencing. To do so, a number of jobs equal to the number of parameters will be passed to the given evaluator. As approximation the finite differencing with epsilon set via the class constructor will be used.
- Parameters
point (numpy array) – The point in parameter space to calculate the jacobi matrix at
evaluator (Evaluator) – the evaluator to use
target (Evaluation) – the target of the calibration, needed only to convert all evaluations to the correct format
result (Result) – The result object to log to
- Returns
the jacobi matrix, and the evaluation at ‘point’
- Return type
tuple (numpy array, Evaluation)
- measurementToNumpyArrayConverter(evaluations, target)
Helper function to convert an array of Evaluation. Each evaluation will be converted and interpolated using it’s getNumpyArrayLike method. None-values or Errors will be converted to None.
- Parameters
evaluations (list of Evaluations) – the evaluations to convert
target – Evaluation describing the format/time steps
each evaluation should be converted/interpolated to :type target: Evaluation :return: the results of the covnertions :rtype: list of numpy arrays
- abstract run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
UGParameterEstimator.optimizers.scipyOptimizers module
- class UGParameterEstimator.optimizers.scipyOptimizers.ScipyMinimizeOptimizer(parametermanager, opt_method='L-BFGS-B', epsilon=0.0001, callback_root=False, callback_scaling=1, differencing=Differencing.forward)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()
- class UGParameterEstimator.optimizers.scipyOptimizers.ScipyNonlinearLeastSquaresOptimizer(parametermanager: ParameterManager, epsilon=0.001, differencing=Differencing.forward)
Bases:
Optimizer- run(evaluator, initial_parameters, target, result=<UGParameterEstimator.dataanalysis.result.Result object>)
Runs this optimizer.
- Parameters
evaluator (Evaluator) – the evaluator to use for each Evaluation needed
initial_parameters (numpy array) – The initial parameters to start the optimization from
target (Evaluation) – The target of the calibration
result (Result, optional) – Results object to write metadata and iterations to, defaults to Result()