Base class for statistics algorithms. More...

#include <vtkStatisticsAlgorithm.h>

Inheritance diagram for vtkStatisticsAlgorithm:

Collaboration diagram for vtkStatisticsAlgorithm:

Classes
class	AssessFunctor
Public Types
typedef vtkTableAlgorithm	Superclass

enum	InputPorts { INPUT_DATA = 0, LEARN_PARAMETERS = 1, INPUT_MODEL = 2 }

enum	OutputIndices { OUTPUT_DATA = 0, OUTPUT_MODEL = 1, ASSESSMENT = 2, OUTPUT_TEST = 2 }
Public Member Functions
virtual int	IsA (const char *type)
vtkStatisticsAlgorithm *	NewInstance () const
void	PrintSelf (ostream &os, vtkIndent indent)
virtual void	SetColumnStatus (const char *namCol, int status)
virtual void	ResetAllColumnStates ()
virtual int	RequestSelectedColumns ()
virtual void	ResetRequests ()
virtual vtkIdType	GetNumberOfRequests ()
virtual vtkIdType	GetNumberOfColumnsForRequest (vtkIdType request)
void	AddColumn (const char *namCol)
void	AddColumnPair (const char namColX, const char namColY)

virtual void	SetLearnOptionParameterConnection (vtkAlgorithmOutput *params)

virtual void	SetLearnOptionParameters (vtkDataObject *params)

virtual void	SetInputModelConnection (vtkAlgorithmOutput *model)

virtual void	SetInputModel (vtkDataObject *model)

virtual void	SetLearnOption (bool)
virtual bool	GetLearnOption ()

virtual void	SetDeriveOption (bool)
virtual bool	GetDeriveOption ()

virtual void	SetAssessOption (bool)
virtual bool	GetAssessOption ()

virtual void	SetTestOption (bool)
virtual bool	GetTestOption ()

virtual void	SetNumberOfPrimaryTables (vtkIdType)
virtual vtkIdType	GetNumberOfPrimaryTables ()

virtual void	SetAssessNames (vtkStringArray *)
virtual vtkStringArray *	GetAssessNames ()

virtual const char *	GetColumnForRequest (vtkIdType r, vtkIdType c)
virtual int	GetColumnForRequest (vtkIdType r, vtkIdType c, vtkStdString &columnName)

virtual bool	SetParameter (const char *parameter, int index, vtkVariant value)

virtual void	Aggregate (vtkDataObjectCollection , vtkMultiBlockDataSet )=0
Static Public Member Functions
static int	IsTypeOf (const char *type)
static vtkStatisticsAlgorithm *	SafeDownCast (vtkObjectBase *o)
Protected Member Functions
virtual vtkObjectBase *	NewInstanceInternal () const
	vtkStatisticsAlgorithm ()
	~vtkStatisticsAlgorithm ()
virtual int	FillInputPortInformation (int port, vtkInformation *info)
virtual int	FillOutputPortInformation (int port, vtkInformation *info)
virtual int	RequestData (vtkInformation , vtkInformationVector , vtkInformationVector )
virtual void	Derive (vtkMultiBlockDataSet *)=0

virtual void	Learn (vtkTable , vtkTable , vtkMultiBlockDataSet *)=0

virtual void	Assess (vtkTable , vtkMultiBlockDataSet , vtkTable *)=0

void	Assess (vtkTable , vtkMultiBlockDataSet , vtkTable *, int)

virtual void	Test (vtkTable , vtkMultiBlockDataSet , vtkTable *)=0

virtual void	SelectAssessFunctor (vtkTable outData, vtkDataObject inMeta, vtkStringArray rowNames, AssessFunctor &dfunc)=0
Protected Attributes
int	NumberOfPrimaryTables
bool	LearnOption
bool	DeriveOption
bool	AssessOption
bool	TestOption
vtkStringArray *	AssessNames
vtkStatisticsAlgorithmPrivate *	Internals

Detailed Description

Base class for statistics algorithms.

All statistics algorithms can conceptually be operated with several operations: Learn: given an input data set, calculate a minimal statistical model (e.g., sums, raw moments, joint probabilities). Derive: given an input minimal statistical model, derive the full model (e.g., descriptive statistics, quantiles, correlations, conditional probabilities). NB: It may be, or not be, a problem that a full model was not derived. For instance, when doing parallel calculations, one only wants to derive the full model after all partial calculations have completed. On the other hand, one can also directly provide a full model, that was previously calculated or guessed, and not derive a new one. Assess: given an input data set, input statistics, and some form of threshold, assess a subset of the data set. Test: perform at least one statistical test. Therefore, a vtkStatisticsAlgorithm has the following ports 3 optional input ports: Data (vtkTable) Parameters to the learn operation (vtkTable) Input model (vtkMultiBlockDataSet) 3 output ports: Data (input annotated with assessments when the Assess operation is ON). Output model (identical to the the input model when Learn operation is OFF). Output of statistical tests. Some engines do not offer such tests yet, in which case this output will always be empty even when the Test operation is ON.

Thanks:: Thanks to Philippe Pebay and David Thompson from Sandia National Laboratories for implementing this class. Updated by Philippe Pebay, Kitware SAS 2012

Examples:: vtkStatisticsAlgorithm (Examples)

Tests:: vtkStatisticsAlgorithm (Tests)

Definition at line 75 of file vtkStatisticsAlgorithm.h.

Member Typedef Documentation

typedef vtkTableAlgorithm vtkStatisticsAlgorithm::Superclass

Member Enumeration Documentation

enum vtkStatisticsAlgorithm::InputPorts

enumeration values to specify input port types

Enumerator:

INPUT_DATA	Port 0 is for learn data.
LEARN_PARAMETERS	Port 1 is for learn parameters (initial guesses, etc.)
INPUT_MODEL	Port 2 is for a priori models.

Definition at line 84 of file vtkStatisticsAlgorithm.h.

enum vtkStatisticsAlgorithm::OutputIndices

enumeration values to specify output port types

Enumerator:

OUTPUT_DATA	Output 0 mirrors the input data, plus optional assessment columns.
OUTPUT_MODEL	Output 1 contains any generated model.
ASSESSMENT	This is an old, deprecated name for OUTPUT_TEST.
OUTPUT_TEST	Output 2 contains result of statistical test(s)

Reimplemented in vtkPairwiseExtractHistogram2D, and vtkExtractHistogram2D.

Definition at line 94 of file vtkStatisticsAlgorithm.h.

Constructor & Destructor Documentation

vtkStatisticsAlgorithm::vtkStatisticsAlgorithm ( ) [protected]

vtkStatisticsAlgorithm::~vtkStatisticsAlgorithm ( ) [protected]

Member Function Documentation

static int vtkStatisticsAlgorithm::IsTypeOf ( const char * name ) [static]

Return 1 if this class type is the same type of (or a subclass of) the named class. Returns 0 otherwise. This method works in combination with vtkTypeMacro found in vtkSetGet.h.

Reimplemented from vtkTableAlgorithm.

Reimplemented in vtkKMeansStatistics, vtkMultiCorrelativeStatistics, vtkDescriptiveStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, vtkOrderStatistics, vtkExtractHistogram2D, vtkPCAStatistics, vtkPairwiseExtractHistogram2D, vtkPOrderStatistics, vtkPContingencyStatistics, vtkPCAStatisticsGnuR, vtkAutoCorrelativeStatistics, vtkPPairwiseExtractHistogram2D, vtkContingencyStatisticsGnuR, vtkCorrelativeStatisticsGnuR, vtkHighestDensityRegionsStatistics, vtkDescriptiveStatisticsGnuR, vtkPExtractHistogram2D, vtkPCorrelativeStatistics, vtkPDescriptiveStatistics, vtkPMultiCorrelativeStatistics, vtkPPCAStatistics, vtkPKMeansStatistics, and vtkPAutoCorrelativeStatistics.

virtual int vtkStatisticsAlgorithm::IsA ( const char * name ) [virtual]

Return 1 if this class is the same type of (or a subclass of) the named class. Returns 0 otherwise. This method works in combination with vtkTypeMacro found in vtkSetGet.h.

Reimplemented from vtkTableAlgorithm.

Reimplemented in vtkKMeansStatistics, vtkMultiCorrelativeStatistics, vtkDescriptiveStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, vtkOrderStatistics, vtkExtractHistogram2D, vtkPCAStatistics, vtkPairwiseExtractHistogram2D, vtkPOrderStatistics, vtkPContingencyStatistics, vtkPCAStatisticsGnuR, vtkAutoCorrelativeStatistics, vtkPPairwiseExtractHistogram2D, vtkContingencyStatisticsGnuR, vtkCorrelativeStatisticsGnuR, vtkHighestDensityRegionsStatistics, vtkDescriptiveStatisticsGnuR, vtkPExtractHistogram2D, vtkPCorrelativeStatistics, vtkPDescriptiveStatistics, vtkPMultiCorrelativeStatistics, vtkPPCAStatistics, vtkPKMeansStatistics, and vtkPAutoCorrelativeStatistics.

static vtkStatisticsAlgorithm* vtkStatisticsAlgorithm::SafeDownCast ( vtkObjectBase * o ) [static]

virtual vtkObjectBase* vtkStatisticsAlgorithm::NewInstanceInternal ( ) const [protected, virtual]

vtkStatisticsAlgorithm* vtkStatisticsAlgorithm::NewInstance ( ) const

void vtkStatisticsAlgorithm::PrintSelf	(	ostream &	os,
		vtkIndent	indent
	)		`[virtual]`

Methods invoked by print to print information about the object including superclasses. Typically not called by the user (use Print() instead) but used in the hierarchical print process to combine the output of several classes.

Reimplemented from vtkTableAlgorithm.

Reimplemented in vtkKMeansStatistics, vtkMultiCorrelativeStatistics, vtkDescriptiveStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, vtkOrderStatistics, vtkExtractHistogram2D, vtkPCAStatistics, vtkPairwiseExtractHistogram2D, vtkPOrderStatistics, vtkPContingencyStatistics, vtkPCAStatisticsGnuR, vtkAutoCorrelativeStatistics, vtkPPairwiseExtractHistogram2D, vtkContingencyStatisticsGnuR, vtkCorrelativeStatisticsGnuR, vtkHighestDensityRegionsStatistics, vtkDescriptiveStatisticsGnuR, vtkPExtractHistogram2D, vtkPCorrelativeStatistics, vtkPDescriptiveStatistics, vtkPMultiCorrelativeStatistics, vtkPPCAStatistics, vtkPKMeansStatistics, and vtkPAutoCorrelativeStatistics.

virtual void vtkStatisticsAlgorithm::SetLearnOptionParameterConnection ( vtkAlgorithmOutput * params ) [inline, virtual]

A convenience method for setting learn input parameters (if one is expected or allowed). It is equivalent to calling SetInputConnection( 1, params );

Definition at line 110 of file vtkStatisticsAlgorithm.h.

virtual void vtkStatisticsAlgorithm::SetLearnOptionParameters ( vtkDataObject * params ) [inline, virtual]

A convenience method for setting learn input parameters (if one is expected or allowed). It is equivalent to calling SetInputData( 1, params );

Definition at line 118 of file vtkStatisticsAlgorithm.h.

virtual void vtkStatisticsAlgorithm::SetInputModelConnection ( vtkAlgorithmOutput * model ) [inline, virtual]

A convenience method for setting the input model connection (if one is expected or allowed). It is equivalent to calling SetInputConnection( 2, model );

Definition at line 126 of file vtkStatisticsAlgorithm.h.

virtual void vtkStatisticsAlgorithm::SetInputModel ( vtkDataObject * model ) [inline, virtual]

A convenience method for setting the input model (if one is expected or allowed). It is equivalent to calling SetInputData( 2, model );

Definition at line 133 of file vtkStatisticsAlgorithm.h.

virtual void vtkStatisticsAlgorithm::SetLearnOption ( bool ) [virtual]

Set/Get the Learn operation.

virtual bool vtkStatisticsAlgorithm::GetLearnOption ( ) [virtual]

Set/Get the Learn operation.

virtual void vtkStatisticsAlgorithm::SetDeriveOption ( bool ) [virtual]

Set/Get the Derive operation.

virtual bool vtkStatisticsAlgorithm::GetDeriveOption ( ) [virtual]

Set/Get the Derive operation.

virtual void vtkStatisticsAlgorithm::SetAssessOption ( bool ) [virtual]

Set/Get the Assess operation.

virtual bool vtkStatisticsAlgorithm::GetAssessOption ( ) [virtual]

Set/Get the Assess operation.

virtual void vtkStatisticsAlgorithm::SetTestOption ( bool ) [virtual]

Set/Get the Test operation.

virtual bool vtkStatisticsAlgorithm::GetTestOption ( ) [virtual]

Set/Get the Test operation.

virtual void vtkStatisticsAlgorithm::SetNumberOfPrimaryTables ( vtkIdType ) [virtual]

Set/Get the number of tables in the primary model.

virtual vtkIdType vtkStatisticsAlgorithm::GetNumberOfPrimaryTables ( ) [virtual]

Set/Get the number of tables in the primary model.

virtual void vtkStatisticsAlgorithm::SetAssessNames ( vtkStringArray * ) [virtual]

Set/get assessment names.

virtual vtkStringArray* vtkStatisticsAlgorithm::GetAssessNames ( ) [virtual]

Set/get assessment names.

virtual void vtkStatisticsAlgorithm::SetColumnStatus	(	const char *	namCol,
		int	status
	)		`[virtual]`

Add or remove a column from the current analysis request. Once all the column status values are set, call RequestSelectedColumns() before selecting another set of columns for a different analysis request. The way that columns selections are used varies from algorithm to algorithm. Note: the set of selected columns is maintained in vtkStatisticsAlgorithmPrivate::Buffer until RequestSelectedColumns() is called, at which point the set is appended to vtkStatisticsAlgorithmPrivate::Requests. If there are any columns in vtkStatisticsAlgorithmPrivate::Buffer at the time RequestData() is called, RequestSelectedColumns() will be called and the selection added to the list of requests.

virtual void vtkStatisticsAlgorithm::ResetAllColumnStates ( ) [virtual]

Set the the status of each and every column in the current request to OFF (0).

virtual int vtkStatisticsAlgorithm::RequestSelectedColumns ( ) [virtual]

Use the current column status values to produce a new request for statistics to be produced when RequestData() is called. See SetColumnStatus() for more information.

virtual void vtkStatisticsAlgorithm::ResetRequests ( ) [virtual]

Empty the list of current requests.

virtual vtkIdType vtkStatisticsAlgorithm::GetNumberOfRequests ( ) [virtual]

Return the number of requests. This does not include any request that is in the column-status buffer but for which RequestSelectedColumns() has not yet been called (even though it is possible this request will be honored when the filter is run -- see SetColumnStatus() for more information).

virtual vtkIdType vtkStatisticsAlgorithm::GetNumberOfColumnsForRequest ( vtkIdType request ) [virtual]

Return the number of columns for a given request.

virtual const char* vtkStatisticsAlgorithm::GetColumnForRequest	(	vtkIdType	r,
		vtkIdType	c
	)		`[virtual]`

Provide the name of the c-th column for the r-th request. For the version of this routine that returns an integer, if the request or column does not exist because r or c is out of bounds, this routine returns 0 and the value of columnName is unspecified. Otherwise, it returns 1 and the value of columnName is set. For the version of this routine that returns const char*, if the request or column does not exist because r or c is out of bounds, the routine returns NULL. Otherwise it returns the column name. This version is not thread-safe.

virtual int vtkStatisticsAlgorithm::GetColumnForRequest	(	vtkIdType	r,
		vtkIdType	c,
		vtkStdString &	columnName
	)		`[virtual]`

Provide the name of the c-th column for the r-th request. For the version of this routine that returns an integer, if the request or column does not exist because r or c is out of bounds, this routine returns 0 and the value of columnName is unspecified. Otherwise, it returns 1 and the value of columnName is set. For the version of this routine that returns const char*, if the request or column does not exist because r or c is out of bounds, the routine returns NULL. Otherwise it returns the column name. This version is not thread-safe.

void vtkStatisticsAlgorithm::AddColumn ( const char * namCol )

Convenience method to create a request with a single column name namCol in a single call; this is the preferred method to select columns, ensuring selection consistency (a single column per request). Warning: no name checking is performed on namCol; it is the user's responsibility to use valid column names.

void vtkStatisticsAlgorithm::AddColumnPair	(	const char *	namColX,
		const char *	namColY
	)

Convenience method to create a request with a single column name pair (namColX, namColY) in a single call; this is the preferred method to select columns pairs, ensuring selection consistency (a pair of columns per request). Unlike SetColumnStatus(), you need not call RequestSelectedColumns() after AddColumnPair(). Warning: namColX and namColY are only checked for their validity as strings; no check is made that either are valid column names.

virtual bool vtkStatisticsAlgorithm::SetParameter	(	const char *	parameter,
		int	index,
		vtkVariant	value
	)		`[virtual]`

A convenience method (in particular for access from other applications) to set parameter values of Learn mode. Return true if setting of requested parameter name was excuted, false otherwise. NB: default method (which is sufficient for most statistics algorithms) does not have any Learn parameters to set and always returns false.

Reimplemented in vtkPCAStatistics, vtkKMeansStatistics, and vtkOrderStatistics.

virtual void vtkStatisticsAlgorithm::Aggregate	(	vtkDataObjectCollection *	,
		vtkMultiBlockDataSet *
	)		`[pure virtual]`

Given a collection of models, calculate aggregate model

Implemented in vtkExtractHistogram2D, vtkKMeansStatistics, vtkPairwiseExtractHistogram2D, vtkOrderStatistics, vtkDescriptiveStatistics, vtkMultiCorrelativeStatistics, vtkAutoCorrelativeStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, and vtkHighestDensityRegionsStatistics.

virtual int vtkStatisticsAlgorithm::FillInputPortInformation	(	int	port,
		vtkInformation *	info
	)		`[protected, virtual]`

Fill the input port information objects for this algorithm. This is invoked by the first call to GetInputPortInformation for each port so subclasses can specify what they can handle.

Reimplemented from vtkTableAlgorithm.

Reimplemented in vtkPCAStatistics.

virtual int vtkStatisticsAlgorithm::FillOutputPortInformation	(	int	port,
		vtkInformation *	info
	)		`[protected, virtual]`

Fill the output port information objects for this algorithm. This is invoked by the first call to GetOutputPortInformation for each port so subclasses can specify what they can handle.

Reimplemented from vtkTableAlgorithm.

Reimplemented in vtkExtractHistogram2D, and vtkPairwiseExtractHistogram2D.

virtual int vtkStatisticsAlgorithm::RequestData	(	vtkInformation *	request,
		vtkInformationVector **	inputVector,
		vtkInformationVector *	outputVector
	)		`[protected, virtual]`

This is called by the superclass. This is the method you should override.

Reimplemented from vtkTableAlgorithm.

virtual void vtkStatisticsAlgorithm::Learn	(	vtkTable *	,
		vtkTable *	,
		vtkMultiBlockDataSet *
	)		`[protected, pure virtual]`

Execute the calculations required by the Learn option, given some input Data

Implemented in vtkExtractHistogram2D, vtkKMeansStatistics, vtkPairwiseExtractHistogram2D, vtkOrderStatistics, vtkDescriptiveStatistics, vtkMultiCorrelativeStatistics, vtkAutoCorrelativeStatistics, vtkHighestDensityRegionsStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, vtkPOrderStatistics, vtkPContingencyStatistics, vtkPMultiCorrelativeStatistics, vtkPExtractHistogram2D, vtkPPCAStatistics, vtkPCorrelativeStatistics, vtkPDescriptiveStatistics, and vtkPAutoCorrelativeStatistics.

virtual void vtkStatisticsAlgorithm::Derive ( vtkMultiBlockDataSet * ) [protected, pure virtual]

Execute the calculations required by the Derive option.

Implemented in vtkPCAStatistics, vtkExtractHistogram2D, vtkKMeansStatistics, vtkPairwiseExtractHistogram2D, vtkOrderStatistics, vtkDescriptiveStatistics, vtkMultiCorrelativeStatistics, vtkAutoCorrelativeStatistics, vtkHighestDensityRegionsStatistics, vtkContingencyStatistics, and vtkCorrelativeStatistics.

virtual void vtkStatisticsAlgorithm::Assess	(	vtkTable *	,
		vtkMultiBlockDataSet *	,
		vtkTable *
	)		`[protected, pure virtual]`

Execute the calculations required by the Assess option.

Implemented in vtkPCAStatistics, vtkExtractHistogram2D, vtkKMeansStatistics, vtkPairwiseExtractHistogram2D, vtkOrderStatistics, vtkDescriptiveStatistics, vtkMultiCorrelativeStatistics, vtkAutoCorrelativeStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, and vtkHighestDensityRegionsStatistics.

void vtkStatisticsAlgorithm::Assess	(	vtkTable *	,
		vtkMultiBlockDataSet *	,
		vtkTable *	,
		int
	)		`[protected]`

A convenience implementation for generic assessment with variable number of variables.

virtual void vtkStatisticsAlgorithm::Test	(	vtkTable *	,
		vtkMultiBlockDataSet *	,
		vtkTable *
	)		`[protected, pure virtual]`

Execute the calculations required by the Test option.

Implemented in vtkPCAStatistics, vtkKMeansStatistics, vtkExtractHistogram2D, vtkPairwiseExtractHistogram2D, vtkOrderStatistics, vtkDescriptiveStatistics, vtkMultiCorrelativeStatistics, vtkHighestDensityRegionsStatistics, vtkAutoCorrelativeStatistics, vtkContingencyStatistics, vtkCorrelativeStatistics, vtkPPCAStatistics, vtkPCorrelativeStatistics, and vtkPAutoCorrelativeStatistics.

virtual void vtkStatisticsAlgorithm::SelectAssessFunctor	(	vtkTable *	outData,
		vtkDataObject *	inMeta,
		vtkStringArray *	rowNames,
		AssessFunctor *&	dfunc
	)		`[protected, pure virtual]`