ILDataList Interface Reference

This interface defines a data list for metric calculation, filtering and analysis. More...

Inheritance diagram for ILDataList:

Public Member Functions
void	Add (double data)
void	Adds (LArray< double > data)
int	Count ()
void	Draw (HDC hdc, LPVChartDrawFlags drawFlags, int xGridStep, int yGridStep)
void	FromDoubleList (LArray< double > data)
void	FromIntList (LArray< int > data)
ILDataList *	GaussianFilter (int kSize)
ILDataList *	GradientFilter (LPVEdgeKernel kType, int kSize)
double	Item (int index)
ILDataList *	MagnitudeFilter (LPVEdgeKernel kType, int kSize)
ILDataList *	MaxFilter (int kSize)
LArray< int >	MaxLocalExtrema (double threshold, int num, LPVFindBy findBy, BOOL onePeakMode)
ILDataList *	MeanFilter (int kSize)
ILDataList *	MedianFilter (int kSize)
ILDataList *	MinFilter (int kSize)
LArray< int >	MinLocalExtrema (double threshold, int num, LPVFindBy findBy, BOOL onePeakMode)
LArray< double >	ToDoubleList ()
LArray< int >	ToIntList ()
Public Member Functions inherited from ILObject
ILObject *	Copy ()
LPVErrorCode	Load (LString filename)
void	Reset ()
LPVErrorCode	Save (LString filename)
BOOL	Valid ()

Properties
int	MaxCount [get, set]
	The maximum number of the list. By default, it's 0, means no limitation. If more than the setting number of data is pushed into the list, the earlier ones will be popped.
int	MaxIndex [get]
	The index of the maximum element.
double	Metric (LPVAggregation aggType) [get]
	The aggregation metric of the data list, such as mean, sum, minimum, etc. More...
int	MinIndex [get]
	The index of the minimum element.

Detailed Description

This interface defines a data list for metric calculation, filtering and analysis.

To use this interface, you should create a LDataList object.

Example Code

C++

Code snippets: snippets/ProjectionAndAnalysis.cpp

ILDataListPtr dataList = LDataList::Create();
dataList->Add(1.1);
dataList->Add(2.3);
dataList->Add(4);
double meanValue = dataList->GetMetric(LPVAggregation::LPVAggMax);
double majorValue = dataList->GetMetric(LPVAggregation::LPVAggMajor);

C#

Code snippets: snippets/ProjectionAndAnalysis.cs

LDataList dataList = new LDataList();
dataList.Add(1.1);
dataList.Add(2.3);
dataList.Add(4);
double meanValue = dataList.Metric[LPVAggregation.LPVAggMax];
double majorValue = dataList.Metric[LPVAggregation.LPVAggMajor];

COM

ILDataListPtr dataList = LDataList::Create();
// or, ILDataListPtr dataList(__uuidof(LDataList)), for absence of built-in Create functions (before 2.2.0)
dataList->Add(1.1);
dataList->Add(2.3);
dataList->Add(4);
double meanValue = dataList->Metric[LPVAggregation::LPVAggMax];
double majorValue = dataList->Metric[LPVAggregation::LPVAggMajor];

Member Function Documentation

Add()

void Add

(

double

data

)

Push the supplied data to the end of the list.

Parameters

[in]

data

The input data.

Adds()

void Adds

(

LArray< double >

data

)

Push the collection of data to the end of the list.

Parameters

[in]

data

The input data, should be a collection of double.

Count()

int Count

(

)

Get count of the data in the list.

Draw()

void Draw	(	HDC	hdc,
		LPVChartDrawFlags	drawFlags,
		int	xGridStep,
		int	yGridStep
	)

Draw the data vector as a chart onto provide device context, using the index as x axis and the value as y axis.

Parameters

[in]	hdc	Handle to the device context of the destination window or bitmap.
[in]	drawFlags	Flags to control which part of a data chart should be drawn, see LPVChartDrawFlags for usage.
[in]	xGridStep	Specify the step of label on x axis, 0 means we'll determine it on the fly according to the score data.
[in]	yGridStep	Specify the step of label on y axis, 0 means we'll determine it on the fly according to the score data. This function is currently not available on Linux / MacOS platform.

FromDoubleList()

void FromDoubleList

(

LArray< double >

data

)

Convert the data from a list of double values.

See also

ToDoubleList().

FromIntList()

void FromIntList

(

LArray< int >

data

)

Convert the data from a list of integer values.

See also

ToIntList().

GaussianFilter()

ILDataList* GaussianFilter

(

int

kSize

)

Blurs the data list using a Gaussian filter. The definition of Gaussian kernel is the same as the Image Gaussian Filter

Parameters

[in]

kSize

The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

GradientFilter()

ILDataList* GradientFilter	(	LPVEdgeKernel	kType,
		int	kSize
	)

The "edges" are represented by bigger absolute values. If a kernel size is set to be 1, the kernel degenerated to be \( (-1, 1) \) no mater what kernel type it is.

Parameters

[in]	kType	The edging kernel type
[in]	kSize	The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

Item()

double Item

(

int

index

)

Fetch the data with the given index.

Parameters

[in]

index

The input index, it should be a 0-based number smaller than total count.

Return values

value

Return the data of given index.

MagnitudeFilter()

ILDataList* MagnitudeFilter	(	LPVEdgeKernel	kType,
		int	kSize
	)

Similar as GradientFilter(), it outputs the magnitude of the gradients, aka. the absolute value of the gradient filter.

Parameters

[in]	kType	The edging kernel type
[in]	kSize	The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

MaxFilter()

ILDataList* MaxFilter

(

int

kSize

)

Each output data represents the maximum of the values in the kernel neighborhood, as:

\[ X{i}' = \max_{j=0}^{kSize} X{i+j-kSize/2} \]

Parameters

[in]

kSize

The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

MaxLocalExtrema()

LArray<int> MaxLocalExtrema	(	double	threshold,
		int	num,
		LPVFindBy	findBy,
		BOOL	onePeakMode
	)

Find the local maximum extrema values in the data list.

Parameters

[in]	threshold	The threshold value used to filter noise data, the values smaller than the threshold is not considered during calculation. By default, it's the minimum double value, means no filter is needed.
[in]	num	The expected number of local extrema values. If there are more extrema points, the top ones are picked according to the specified findBy strategy. By default, it's 0 means find all.
[in]	findBy	The searching strategy.
[in]	onePeakMode	Turn it on to enable to one-peak mode, then if there are multiple extrema values in one thresholded region, only the largest one is returned. By default, it's off.

Return values

indexes

Return the extrema indexes ascendingly.

See also

MinLocalExtrema()

MeanFilter()

ILDataList* MeanFilter

(

int

kSize

)

Each output data represents the mean of the values in the kernel neighborhood, as:

\[ X{i}' = \sum_{j=0}^{kSize} X{i+j-kSize/2} \]

Parameters

[in]

kSize

The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

MedianFilter()

ILDataList* MedianFilter

(

int

kSize

)

Each output data represents the median of the values in the kernel neighborhood.

Parameters

[in]

kSize

The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

MinFilter()

ILDataList* MinFilter

(

int

kSize

)

Each output data represents the minimum of the values in the kernel neighborhood, as:

\[ X{i}' = \min_{j=0}^{kSize} X{i+j-kSize/2} \]

Parameters

[in]

kSize

The kernel size used in the filter, should be an odd number, say, 1, 3, 5, 7, etc. By default it's 3. Large kernel size results usually in stronger filtering effect, but also slower.

Return values

result

Return the filtered data list

MinLocalExtrema()

LArray<int> MinLocalExtrema	(	double	threshold,
		int	num,
		LPVFindBy	findBy,
		BOOL	onePeakMode
	)

Find the local minimum extrema values in the data list.

Parameters

[in]	threshold	The threshold value used to filter noise data, the values bigger than the threshold is not considered during calculation. By default, it's the maximum double value, means no filter is needed.
[in]	num	The expected number of local extrema values. If there are more extrema points, the top ones are picked according to the specified findBy strategy. By default, it's 0 means find all.
[in]	findBy	The searching strategy.
[in]	onePeakMode	Turn it on to enable to one-peak mode, then if there are multiple extrema values in one thresholded region, only the smallest one is returned. By default, it's off.

Return values

indexes

Return the extrema indexes ascendingly.

See also

MaxLocalExtrema()

ToDoubleList()

LArray<double> ToDoubleList

(

)

Convert the data to a list of double values.

See also

FromDoubleList().

ToIntList()

LArray<int> ToIntList

(

)

Convert the data to a list of integer values.

See also

FromIntList().

Property Documentation

Metric

double Metric( LPVAggregation aggType)

get

The aggregation metric of the data list, such as mean, sum, minimum, etc.

Parameters

[in]

aggType

The aggregation type

Return values

val	Return the metric value

See also

MinIndex, MaxIndex