ApproxTrainData: Data class for approximation training data, containing several useful bounding box properties etc. More...

Detailed Description

ApproxTrainData: Data class for approximation training data, containing several useful bounding box properties etc.

Author: Daniel Wirtz

Date: 2011-11-02

New in 0.6:: (Daniel Wirtz, 2012-07-18) Moved computation method for approx train data from models.BaseFullModel to here.

New in 0.6:: (Daniel Wirtz, 2011-11-16) Added a new method data.ApproxTrainData.getCombinedData which returns the triples \(x_i,t_i,\mu_i\) as one column matrix.

New in 0.5:: (Daniel Wirtz, 2011-11-02) Added this class.

Todo:: Make xi,ti,mui private fields and update dependents like Center upon re-setting one of those values! (consistency issue)

This class is part of the framework

KerMor - Model Order Reduction using Kernels

Homepage http://www.morepas.org/software/index.html
Documentation http://www.morepas.org/software/kermor/index.html
License KerMor license conditions

Definition at line 18 of file ApproxTrainData.m.

Public Member Functions
	ApproxTrainData (matrix< double > xi,rowvec< double > ti,matrix< double > mui)
	Creates a new approx train data instance. More...

function [ colvec< double > c , integer idx ] =	getClosestToCenter ()
	The sample triple closest to the Center. More...

function satd =	subset (sel)

function [ train , validation , randidx ] =	splitToTrainValidationSets (perc, seed)
	Creates a training and validation data set from this approx train data instance. More...

function [ e , lbl , pt ] =	getErrorsFor (fun)

function	relocate (char new_dir)
	Convenience method. Relocates the xi and fxi FileMatrix instances if present. More...

function	delete ()

function	makeUniqueXi ()

function [ minDist , meanDist , maxDist ] =	getXiDists ()

Static Public Member Functions
static function data.ApproxTrainData atd =	computeFrom (models.BaseFullModel model,dscomponents.ACoreFun f,data.selection.ASelector selector,logical parallel)
	Computes approximation training data for a model, function and selector. More...

static function [ data.ApproxTrainData atd , minScale , maxScale ] =	scaleXiZeroOne (data.ApproxTrainData atd)

Public Attributes
data.FileMatrix	xi = "[]"
	The state space samples \(x_i = x(t_i;\mu_i)\), stored row-wise in a data.FileMatrix instance. More...

rowvec	ti = "[]"
	The time samples \(t_i\). More...

matrix	mui = "[]"
	The parameter samples \(\mu_i\) used computing the parent trajectories of \(x_i\). More...

data.FileMatrix	fxi
	The evaluations of \(f(x_i,t_i,\mu_i)\), stored row-wise in a data.FileMatrix. More...

colvec	Center
	The geometrical center of the datas bounding box. More...

double	xiDia
	The diameter of the state space samples. More...

double	tiDia
	The time span of the time samples. More...

double	muiDia
	The diameter of the parameter space samples. More...

logical	hasTime = false
	Flag that indicates if time samples are present. More...

logical	hasParams = false
	Flag that indicates if param samples are present. More...

integer	tOff = "[]"
	The index of the time entry in the combined \([x;t;\mu]\) vector. More...

integer	muOff = "[]"
	The index of the first parameter entry in the combined \([x;t;\mu]\) vector. More...

struct	Box = "[]"
	A box struct allowing access to the specific min and max values of the samples. Recommended for Debug use only. More...

Public Attributes inherited from handle
	addlistener
	Creates a listener for the specified event and assigns a callback function to execute when the event occurs. More...

	notify
	Broadcast a notice that a specific event is occurring on a specified handle object or array of handle objects. More...

	delete
	Handle object destructor method that is called when the object's lifecycle ends. More...

	disp
	Handle object disp method which is called by the display method. See the MATLAB disp function. More...

	display
	Handle object display method called when MATLAB software interprets an expression returning a handle object that is not terminated by a semicolon. See the MATLAB display function. More...

	findobj
	Finds objects matching the specified conditions from the input array of handle objects. More...

	findprop
	Returns a meta.property objects associated with the specified property name. More...

	fields
	Returns a cell array of string containing the names of public properties. More...

	fieldnames
	Returns a cell array of string containing the names of public properties. See the MATLAB fieldnames function. More...

	isvalid
	Returns a logical array in which elements are true if the corresponding elements in the input array are valid handles. This method is Sealed so you cannot override it in a handle subclass. More...

	eq
	Relational functions example. See details for more information. More...

	transpose
	Transposes the elements of the handle object array. More...

	permute
	Rearranges the dimensions of the handle object array. See the MATLAB permute function. More...

	reshape
	hanges the dimensions of the handle object array to the specified dimensions. See the MATLAB reshape function. More...

	sort
	ort the handle objects in any array in ascending or descending order. More...

Constructor & Destructor Documentation

data.ApproxTrainData.ApproxTrainData	(	matrix< double >	xi,
		rowvec< double >	ti,
		matrix< double >	mui
	)

Creates a new approx train data instance.

Leave \(t_i\) and/or \(\mu_i\) empty if no time or parameter training data is present.

Assign local variables

Parameters

xi	The state space samples \(x_i\)
ti	The time instances \(t_i\)
mui	The parameter samples \(\mu_i\)

Required fields of xi:

Definition at line 218 of file ApproxTrainData.m.

References Box, Center, Utils.getBoundingBox(), hasParams, hasTime, mui, muiDia, muOff, ti, tiDia, tOff, xi, and xiDia.

Referenced by models.BaseFullModel.off4_genApproximationTrainData().

Here is the call graph for this function:

Here is the caller graph for this function:

Member Function Documentation

function data.ApproxTrainData atd = data.ApproxTrainData.computeFrom	(	models.BaseFullModel	model,
		dscomponents.ACoreFun	f,
		data.selection.ASelector	selector,
		logical	parallel
	)

static

Computes approximation training data for a model, function and selector.

This method is implemented here as both the original offline phase 4 uses this method but also the DEIM error estimator needs the same routine for the MatrixDEIM approximation.

See Also: models.BaseFullModel error.DEIMEstimator

Parameters

model	The full model instance
f	The model's core function
selector	The training data selector
parallel	Flag to set if computation should be done in parallel (MatlabPool) Default: false

Required fields of selector:

Required fields of model:

Required fields of f:

Generated fields of atd:

Definition at line 456 of file ApproxTrainData.m.

References KerMor.App(), data.FileMatrix.bCols, models.BaseFullModel.Data, data.FileData.DataDirectory, dscomponents.ACoreFun.evaluateMulti(), dscomponents.ACoreFun.fDim, data.FileMatrix.FileMatrix(), fxi, mui, data.FileMatrix.n, data.FileMatrix.nBlocks, models.BaseFullModel.ProjectApproxTrainingData, data.selection.ASelector.selectTrainingData(), t, and xi.

Here is the call graph for this function:

function data.ApproxTrainData.delete ( )

Definition at line 382 of file ApproxTrainData.m.

References fxi, and xi.

function [ colvec< double > c , integer idx ] = data.ApproxTrainData.getClosestToCenter ( )

The sample triple closest to the Center.

Return values

c	A column vector composed as \([x; t; mu]\)
idx	The index in xi,ti,mui samples

Definition at line 275 of file ApproxTrainData.m.

References Center, mui, ti, and xi.

function [e , lbl , pt ] = data.ApproxTrainData.getErrorsFor ( fun )

Definition at line 325 of file ApproxTrainData.m.

References PrintTable.Caption, fxi, k, Norm.L1(), Norm.L2(), t, data.ABlockedData.toMemoryMatrix(), and xi.

Here is the call graph for this function:

function [minDist , meanDist , maxDist ] = data.ApproxTrainData.getXiDists ( )

Definition at line 411 of file ApproxTrainData.m.

References data.ABlockedData.toMemoryMatrix(), and xi.

Here is the call graph for this function:

function data.ApproxTrainData.makeUniqueXi ( )

Definition at line 388 of file ApproxTrainData.m.

References KerMor.App(), fxi, mui, data.FileMatrix.spawnWithContent(), t, ti, data.ABlockedData.toMemoryMatrix(), and xi.

Here is the call graph for this function:

function data.ApproxTrainData.relocate ( char new_dir )

Convenience method. Relocates the xi and fxi FileMatrix instances if present.

New in 0.7:: (Daniel Wirtz, 2013-05-28) Added this method.

Parameters

new_dir The new directory

Definition at line 364 of file ApproxTrainData.m.

References fxi, data.FileMatrix.relocate(), and xi.

Referenced by data.ModelData.relocate().

Here is the call graph for this function:

Here is the caller graph for this function:

static function [data.ApproxTrainDataatd , minScale , maxScale ] = data.ApproxTrainData.scaleXiZeroOne ( data.ApproxTrainData atd )

static

Definition at line 562 of file ApproxTrainData.m.

References xi.

function [ train , validation , randidx ] = data.ApproxTrainData.splitToTrainValidationSets	(	perc,
		seed
	)

Creates a training and validation data set from this approx train data instance.

Definition at line 309 of file ApproxTrainData.m.

References subset(), and xi.

Here is the call graph for this function:

function satd = data.ApproxTrainData.subset ( sel )

Definition at line 291 of file ApproxTrainData.m.

References fxi, mui, ti, and xi.

Referenced by splitToTrainValidationSets().

Here is the caller graph for this function:

Member Data Documentation

data.ApproxTrainData.Box = "[]"

A box struct allowing access to the specific min and max values of the samples. Recommended for Debug use only.

Note: This property has non-standard access specifiers: SetAccess = Private, GetAccess = Public; Matlab documentation of property attributes.
Default: "[]"

Definition at line 201 of file ApproxTrainData.m.

Referenced by ApproxTrainData().

data.ApproxTrainData.Center

The geometrical center of the datas bounding box.

A column vector composed as \([x; t; mu]\)

Note: This property has non-standard access specifiers: SetAccess = Private, GetAccess = Public; Matlab documentation of property attributes.

Definition at line 94 of file ApproxTrainData.m.

Referenced by ApproxTrainData(), and getClosestToCenter().

data.ApproxTrainData.fxi

The evaluations of \(f(x_i,t_i,\mu_i)\), stored row-wise in a data.FileMatrix.

Note: This property has custom functionality when its value is changed.

Definition at line 82 of file ApproxTrainData.m.

Referenced by approx.DEIM.approximateSystemFunction(), approx.algorithms.ABase.computeApproximation(), testing.DEIM.computeDEIMErrors(), computeFrom(), delete(), approx.algorithms.ABase.getError(), getErrorsFor(), makeUniqueXi(), models.BaseFullModel.off5_computeApproximation(), relocate(), approx.algorithms.VKOGA.startAdaptiveExtension(), subset(), and approx.algorithms.Componentwise.templateComputeApproximation().

data.ApproxTrainData.hasParams = false

Flag that indicates if param samples are present.

Default:: false

Note: This property has non-standard access specifiers: SetAccess = Private, GetAccess = Public; Matlab documentation of property attributes.
Default: false

Definition at line 157 of file ApproxTrainData.m.

Referenced by ApproxTrainData(), approx.algorithms.AAdaptiveBase.extendExpansion(), kernels.config.RBFConfig.getDists(), kernels.config.ExpansionConfig.guessGammaConfigs(), and kernels.ParamTimeKernelExpansion.setCentersFromATD().

data.ApproxTrainData.hasTime = false

Flag that indicates if time samples are present.

Default:: false

Note: This property has non-standard access specifiers: SetAccess = Private, GetAccess = Public; Matlab documentation of property attributes.
Default: false

Definition at line 144 of file ApproxTrainData.m.

Referenced by ApproxTrainData(), approx.algorithms.AAdaptiveBase.extendExpansion(), kernels.config.RBFConfig.getDists(), kernels.config.ExpansionConfig.guessGammaConfigs(), and kernels.ParamTimeKernelExpansion.setCentersFromATD().

data.ApproxTrainData.mui = "[]"

The parameter samples \(\mu_i\) used computing the parent trajectories of \(x_i\).

Default: []

Definition at line 71 of file ApproxTrainData.m.

Referenced by ApproxTrainData(), testing.DEIM.computeDEIMErrors(), computeFrom(), approx.algorithms.AAdaptiveBase.extendExpansion(), getClosestToCenter(), approx.algorithms.ABase.getError(), makeUniqueXi(), FunVis2D>rangesFromATD(), kernels.ParamTimeKernelExpansion.setCentersFromATD(), approx.algorithms.VKOGA.startAdaptiveExtension(), subset(), and approx.algorithms.Componentwise.templateComputeApproximation().