Properties with level "Optional"

Property level optional

Member approx.algorithms.AAdaptiveBase.InitialCenter

The first center to choose should not greatly influence the algorithm outcome. The default maxfx is related the closest to the greedy-type approach.

Member approx.algorithms.ABase.UsefScaling

This option makes sense when using univariate rotation-invariant kernels as different dimensions might have different scales

Member data.selection.TimeSelector.Seed

The seed for the random number generator.

Member DPCMObject.WorkspaceVariableName

This property can be set by any methods that perform computations that use the DPCM system a-priori. Setting this value increases the readability of the DPCM output.

This property can be set by any methods that perform computations that use the DPCM system a-priori. Setting this value increases the readability of the DPCM output.

Member dscomponents.ACoreFun.CustomProjection

If the structure of the underlying function allows more efficient projection set this flag to true in the subclass constructor.

Member dscomponents.ACoreFun.JSparsityPattern

Some ODE solvers can work more efficiently if a sparsity pattern for the jacobian matrix of the core function can be provided.

Member error.lipfun.ImprovedLocalSecantLipschitz.UseCachedSecants

Member general.interpolation.KernelInterpol.UseNewtonBasis

This can greatly improve numerical stability of computation, but possibly takes more time to compute.

Member general.regression.BaseScalarSVR.AlphaRelMinValue

The threshold of 16 magnitudes below the maximum coefficient should be small enough to not have any bad influence but improve performance.

Member general.regression.KernelLS.MaxStraightInvDim

Direct inversion is more precise for small sizes.

Member general.regression.ScalarEpsSVR_SMO.NNk

A higher number of neighbors might

Member general.regression.ScalarEpsSVR_SMO.Vis

Member kernels.BellFunction.NewtonTolerance

The default precision is sufficient for all cases encountered so far.

Member kernels.CombinationKernel.CombOp

Default combination method is pointwise multiplication.

Member kernels.KernelExpansion.Base

A custom base for which the coefficients are set.

Member kernels.ParamTimeKernelExpansion.StateNablaCombinationFun

Adjust this whenever the SubKernelCombinationFun is changed.

Member kernels.ParamTimeKernelExpansion.SubKernelCombinationFun

Standard kernel combination is pointwise multiplication.

Member models.BaseFirstOrderSystem.A

Member models.BaseFirstOrderSystem.B

Member models.BaseFirstOrderSystem.C

Member models.BaseFirstOrderSystem.Inputs

Member models.BaseFirstOrderSystem.Params

Member models.BaseFullModel.AutoSave

This setting just decreases the chance of data loss due to intermediate saves.

Member models.BaseFullModel.EnableTrajectoryCaching

Caching can speedup computations but may require extra disk space.

Member models.BaseFullModel.ErrorEstimator

Using error estimators with reduced models can facilitate reduction quality measurement and can also be needed by some adaptive algorithms.

Member models.BaseFullModel.postApproximationTrainingCallback

A callback that can be set to execute user defined code after training the approximation.

Member models.BaseFullModel.preApproximationTrainingCallback

A callback that can be set to execute user defined code before training the approximation.

Member models.BaseFullModel.SaveTag

Member models.BaseFullModel.TrainingInputs

Set subindices to restrict offline generation phase to a subset of inputs.

Member models.BaseFullModel.TrainingParams

Set subindices to restrict offline generation phase to a subset of parameters.

Member models.BaseModel.G

Member models.BaseModel.Name

Member models.BaseModel.RealTimePlotting

Additionally displays the system's plot during simulations.

Member models.BaseModel.RealTimePlottingMinPause

Changes pause length between timesteps

Member models.beam.DynLinTimoshenkoModel.BeamLineWidth

Visualization setting

Member models.beam.DynLinTimoshenkoModel.NumColors

Member models.beam.DynLinTimoshenkoModel.PlotFactor

Scales the displacement

Member sampling.RandomSampler.Seed

Set to empty for new choices on every run.

Member solvers.BaseSolver.RealTimeMode

This setting is model dependent

Member spacereduction.PODGreedy.InitialSpace

Can be useful if a specific choice for initial spaces is available.