org.simbrain.gauge.core
Class AddData

java.lang.Object
  extended by org.simbrain.gauge.core.AddData

public class AddData
extends Object

AddData is a set of functions for adding new datapoints to existing datasets, when a projection is being used in real-time. These methods will generally be fast as compared with, for example, re-running the Sammon map each time a new point is added.


Constructor Summary
AddData()
           
 
Method Summary
static double[] coordinate(int i, int j, double[] hiPoint)
          Coordinate project new points.
static double[] nnSubspace(Dataset upstairs, Dataset downstairs, double[] hiPoint)
          Adds new datapoint to subspace spanned by nearest-neighbors.
static double[] triangulate(Dataset upstairs, Dataset downstairs, double[] hiPoint)
          Adds a new datapoint which preserves distances to nearest neighbors.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

AddData

public AddData()
Method Detail

coordinate

public static double[] coordinate(int i,
                                  int j,
                                  double[] hiPoint)
Coordinate project new points.

Parameters:
i - first low-d coordinate for projection
j - second low-d coordinate for projection
hiPoint - new point upstairs
Returns:
projected point downstairs

triangulate

public static double[] triangulate(Dataset upstairs,
                                   Dataset downstairs,
                                   double[] hiPoint)
Adds a new datapoint which preserves distances to nearest neighbors.

Parameters:
upstairs - reference to upstairs dataset
downstairs - reference to downstairs dataset
hiPoint - new hi-d point to be added
Returns:
low-d point

nnSubspace

public static double[] nnSubspace(Dataset upstairs,
                                  Dataset downstairs,
                                  double[] hiPoint)
Adds new datapoint to subspace spanned by nearest-neighbors.

Parameters:
upstairs - reference to upstairs dataset
downstairs - reference to downstairs dataset
hiPoint - new hi-d point to be added
Returns:
low-d point