F_VN_CreateLdaTransformViaComponentNum
Create an LDA-based feature transform from the provided data where the number of components to be used is predefined. The number of samples must be >= the number of features and the number of classes must be >= 2. The initial reference count is set to one if a new model is created and kept, otherwise. (expert function)
Syntax
Definition:
FUNCTION F_VN_CreateLdaTransformViaComponentNum : HRESULT
VAR_INPUT
ipMlModel : Reference To ITcVnMlModel;
ipSamples : ITcUnknown;
ipClasses : ITcVnContainer;
nComponentNum : UDINT;
hrPrev : HRESULT;
END_VAR
Inputs
Name |
Type |
Description |
---|---|---|
ipMlModel |
Reference To ITcVnMlModel |
Returns the created feature transform (Non-zero interface pointers are reused.) |
ipSamples |
Container holding a batch of input samples (ContainerType_Vector_Vector_REAL or ContainerType_Vector_Vector_LREAL) | |
ipClasses |
Class labels corresponding to the input samples (ContainerType_Vector_DINT) | |
nComponentNum |
UDINT |
Number of components (nComponentNum must be smaller than the number of classes. A value of 0 results in the automatic computation of the maximum number of components.) |
hrPrev |
HRESULT indicating the result of previous operations (If SUCCEEDED(hrPrev) equals false, no operation is executed.) |
Further information
The function F_VN_CreateLdaTransformViaComponentNum
is a variant of F_VN_CreateLdaTransform. It contains the additional parameter nComponentNum
to specify the number of components for the feature reduction in advance. This allows the level of data compression to be set.
Parameter
Model
The created model is returned in the interface pointer ipMlModel
.
Samples
All samples are sent to the model in a container via ipSamples
.
Classes
The class assignments of all samples in a container are given to the model via ipClasses
.
Number of components
nComponentNum
components are used for feature reduction. At 0, the maximum number of components is automatically calculated.
Application
For example, an LDA model with 25 components can be created like this:
hr := F_VN_CreateLdaTransformViaComponentNum(
ipMlModel := ipMlModel,
ipSamples := ipSamples,
ipClasses := ipClasses,
nComponentNum := 25,
hrPrev := hr);
Related functions
Required License
TC3 Vision Machine Learning
System Requirements
Development environment | Target platform | PLC libraries to include |
---|---|---|
TwinCAT V3.1.4024.54 or later | PC or CX (x64) with PL50, e.g. Intel 4-core Atom CPU | Tc3_Vision |