SharpBCI.AggregatePredictionPipeable Class Reference

A pipeable which aggregates received EEGEvents into an array based on the reported EEGEvent timestamp and then uses an IPredictor<EEGEvent[]> to train/classify on them. More...

Inheritance diagram for SharpBCI.AggregatePredictionPipeable:

Public Member Functions
	AggregatePredictionPipeable (int channels, int k, double thresholdProb, EEGDataType[] types)
	AggregatePredictionPipeable (int channels, int k, double thresholdProb, object[] typeNames)
void	ClearTrainingData ()
	Clears all training data stored within the predictor Essentially a reset of the entire prediction system Use between changing environments and/or participants. More...
void	StartTraining (int id)
	Start training predictor on the EEG data from now on Should be paired w/ a StopTraining(id) call. More...
void	StopTraining (int id)
	Stop training predictor on the EEG data from now on Should be paired w/ a StopTraining(id) call. More...
Public Member Functions inherited from SharpBCI.Pipeable
void	SetInput (BlockingCollection< object > newInput)
	Set the input on this IPipeable to param input Should only be used internally by Connect. More...
void	Connect (IPipeable other)
	Connect the input of other to our output Note: mirror == false using this override. More...
void	Connect (IPipeable other, bool mirror)
	Connect the input of other to our output and allow for control of mirror data. More...
virtual void	Start (TaskFactory taskFactory, CancellationTokenSource cts)
	Actually start doing work (i.e., promise to eventually start pushing data to connected pipeables) More...
virtual void	Stop ()
	Require this IPipeable to stop, blocking until actually stopped. More...

Public Attributes
const int	ID_PREDICT = 0
	Constant used to indicate when the encapsulated predictor is actively predicting. More...
const int	NO_PREDICTION = -1
	Constant used to indicate when the encapsulated predictor has no valid prediction. More...
Public Attributes inherited from SharpBCI.Pipeable
const int	DEFAULT_BUFFER_SIZE = 1000

Protected Member Functions
override bool	Process (object item)
	Collects incoming EEGEvents (one per type), and stores them in buffer. More...
Protected Member Functions inherited from SharpBCI.Pipeable
void	Add (object item)

Detailed Description

A pipeable which aggregates received EEGEvents into an array based on the reported EEGEvent timestamp and then uses an IPredictor<EEGEvent[]> to train/classify on them.

A TrainedEvent will be pushed on to the next stage.

See Also

EEGEvent

TrainedEvent

Constructor & Destructor Documentation

SharpBCI.AggregatePredictionPipeable.AggregatePredictionPipeable	(	int	channels,
		int	k,
		double	thresholdProb,
		EEGDataType[]	types
	)

SharpBCI.AggregatePredictionPipeable.AggregatePredictionPipeable	(	int	channels,
		int	k,
		double	thresholdProb,
		object[]	typeNames
	)

Member Function Documentation

void SharpBCI.AggregatePredictionPipeable.ClearTrainingData

(

)

Clears all training data stored within the predictor Essentially a reset of the entire prediction system Use between changing environments and/or participants.

Implements SharpBCI.IPredictorPipeable.

override bool SharpBCI.AggregatePredictionPipeable.Process ( object  item )

protectedvirtual

Collects incoming EEGEvents (one per type), and stores them in buffer.

If the incoming event's timestamp is new, attempt to send the buffer onto the IPredictor<EEGEvent[]> If the current trainingId is non-zero, the data will be used for training on the trainingId label If the current trainingId is zero, the data will be predicted on

See Also

Pipeable

Implements SharpBCI.Pipeable.

void SharpBCI.AggregatePredictionPipeable.StartTraining

(

int

id

)

Start training predictor on the EEG data from now on Should be paired w/ a StopTraining(id) call.

Parameters

id	- a unique non-negative non-zero integer which identifies this trained event

Implements SharpBCI.IPredictorPipeable.

void SharpBCI.AggregatePredictionPipeable.StopTraining

(

int

id

)

Stop training predictor on the EEG data from now on Should be paired w/ a StopTraining(id) call.

Parameters

id	- a unique non-negative non-zero integer which identifies this trained event

Implements SharpBCI.IPredictorPipeable.

Member Data Documentation

const int SharpBCI.AggregatePredictionPipeable.ID_PREDICT = 0

Constant used to indicate when the encapsulated predictor is actively predicting.

const int SharpBCI.AggregatePredictionPipeable.NO_PREDICTION = -1

Constant used to indicate when the encapsulated predictor has no valid prediction.

---

The documentation for this class was generated from the following file:

• /home/travis/build/berkona/SharpBCI/SharpBCI/AggregatePredictors.cs