classPiiSampleSetCollector
#include <PiiSampleSetCollector.h>
A learning algorithm that just collects all incoming data into a sample set.
Inherits PiiLearningAlgorithm< SampleSet >
Description
Public types
|
typedef PiiSampleSet::Traits< SampleSet
>::ConstFeatureIterator
|
Constructors and destructor
|
Creates a new PiiSampleSetCollector that creates the sample, label and weight sets automatically and takes care of deleting them. |
|
|
Creates a new sample set collector that appends samples to the given sample set, label list and weight list. |
|
|
Creates a new sample set collector that appends samples to the given sample set and label list. |
|
|
(
Creates a new sample set collector that appends samples to the given sample set. |
|
|
Destroys this sample set collector. |
Public member functions
|
int
|
( )
Returns the batch size. |
|
virtual PiiClassification::LearnerCapabilities
|
( )
Returns |
|
QVector<
double > *
|
( )
Returns the class labels. |
|
bool
|
( )
Returns |
|
bool
|
( )
Returns |
|
virtual bool
|
( )
Returns |
|
int
|
( )
Returns the number of features in the stored samples. |
|
Returns the way full buffer is handled. |
|
|
virtual void
|
Replaces all collected samples and labels with the given new ones. |
|
virtual double
|
(
Appends sample to the sample set. |
|
void
|
(
Resizes the collected sample set and associated labels and weight so that at most samples entries are be retained. |
|
int
|
( )
Returns the number of samples currently in the sample set. |
|
SampleSet *
|
( )
Returns the sample set. |
|
QVector<
double > *
|
( )
Returns the sample weights. |
|
void
|
(
Sets the maximum size for the buffer. |
|
void
|
Sets the class labels. |
|
void
|
(
Enables or disables the collecting of class labels. |
|
void
|
(
Enables or disables the collecting of sample weights. |
|
void
|
Sets the strategy of handling a full buffer. |
|
void
|
(
Set the sample set. |
|
void
|
Sets the sample weights. |
Function details
-
PiiSampleSetCollector
()Creates a new PiiSampleSetCollector that creates the sample, label and weight sets automatically and takes care of deleting them.
-
PiiSampleSetCollector
Creates a new sample set collector that appends samples to the given sample set, label list and weight list.
PiiSampleSetCollector doesn't take the ownership of the pointers, and they must remain valid during the lifetime of this object.
-
PiiSampleSetCollector
Creates a new sample set collector that appends samples to the given sample set and label list.
PiiSampleSetCollector doesn't take the ownership of the pointers, and they must remain valid during the lifetime of this object. The weight list will be created internally.
-
PiiSampleSetCollector
(- SampleSet * sampleSet
Creates a new sample set collector that appends samples to the given sample set.
PiiSampleSetCollector doesn't take the ownership of the pointers and it must remain valid during the lifetime of this object. The label and weight lists will be created internally.
-
~PiiSampleSetCollector
()Destroys this sample set collector.
-
int batchSize
()Returns the batch size.
-
virtual PiiClassification::LearnerCapabilities capabilities
()[virtual]Returns
NonSupervisedLearner|OnlineLearner|WeightedLearner;.Reimplemented from PiiLearningAlgorithm.
-
QVector< double > * classLabels
()Returns the class labels.
-
bool collectLabels
()Returns
trueif class labels are being collected,falseotherwise. -
bool collectWeights
()Returns
trueif sample weights are being collected,falseotherwise. -
virtual bool converged
()[virtual]Returns
true.Reimplemented from PiiLearningAlgorithm.
-
int featureCount
()Returns the number of features in the stored samples.
If the sample set is empty, zero will be returned.
-
PiiClassification::FullBufferBehavior fullBufferBehavior
()Returns the way full buffer is handled.
-
virtual void learn
[virtual]Replaces all collected samples and labels with the given new ones.
Reimplemented from PiiLearningAlgorithm.
-
virtual double learnOne
(- ConstFeatureIterator sample
- int length
- double label
- double weight
[virtual]Appends sample to the sample set.
If label collection is enabled, label will be added to the label set. If weight collection is enabled, weight will be added to the weight set. If length doesn't match the current number of features, the current sample set will be cleared first. Returns the index of the just added sample in the sample set.
Reimplemented from PiiLearningAlgorithm.
-
void resize
(- int samples
Resizes the collected sample set and associated labels and weight so that at most samples entries are be retained.
-
int sampleCount
()Returns the number of samples currently in the sample set.
-
SampleSet * samples
()Returns the sample set.
-
QVector< double > * sampleWeights
()Returns the sample weights.
-
void setBatchSize
(- int batchSize
Sets the maximum size for the buffer.
-1 means no limit.
-
void setClassLabels
Sets the class labels.
If the class labels are set to 0, an empty vector will be automatically created.
-
void setCollectLabels
(- bool collectLabels
Enables or disables the collecting of class labels.
If class labels are not collected, the class label list will remain empty. The default value is
true, unless only a sample list is given in the constructor. -
void setCollectWeights
(- bool collectWeights
Enables or disables the collecting of sample weights.
If sample weights are not collected, the weight list will remain empty. The default value is
false, unless a weight list is explicitly given in the constructor. -
void setFullBufferBehavior
Sets the strategy of handling a full buffer.
-
void setSamples
(- SampleSet * samples
Set the sample set.
If the samples are set to 0, a new sample set will be automatically created.
-
void setSampleWeights
Sets the sample weights.
If the weights are set to 0, an empty vector will be automatically created.
Add a note
Not a single note added yet. Be the first, add yours.