hex.util

Class DimensionReductionUtils

java.lang.Object

hex.util.DimensionReductionUtils

public class DimensionReductionUtils
extends java.lang.Object

Created by wendycwong on 2/9/17.

Constructor Summary

Constructors

Constructor and Description
DimensionReductionUtils()

Method Summary

Methods

Modifier and Type	Method and Description
static water.util.TwoDimTable	createScoringHistoryTableDR(java.util.LinkedHashMap<java.lang.String,java.util.ArrayList> scoreTable, java.lang.String tableName, long startTime) Create the scoring history for dimension reduction algorithms like PCA/SVD.
static void	generateIPC(double[] std_deviation, double totVar, double[] vars, double[] prop_var, double[] cum_var) This method will calculate the importance of principal components for PCA/GLRM methods.
static double[][]	transformEigenVectors(DataInfo dinfo, double[][] vEigenIn) This function will tranform the eigenvectors calculated for a matrix T(A) to the ones calculated for matrix A.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DimensionReductionUtils

public DimensionReductionUtils()

Method Detail

generateIPC

public static void generateIPC(double[] std_deviation,
               double totVar,
               double[] vars,
               double[] prop_var,
               double[] cum_var)

This method will calculate the importance of principal components for PCA/GLRM methods.

Parameters:

std_deviation: - array of singular values

totVar: - sum of squared singular values

vars: - array of singular values squared

prop_var: - var[i]/totVar for each i

cum_var: - cumulative sum of var[i]/totVar from index 0 to index i.

createScoringHistoryTableDR

public static water.util.TwoDimTable createScoringHistoryTableDR(java.util.LinkedHashMap<java.lang.String,java.util.ArrayList> scoreTable,
                                                 java.lang.String tableName,
                                                 long startTime)

Create the scoring history for dimension reduction algorithms like PCA/SVD. We do make the following assumptions about your scoring_history. First we assume that you will always have the following field: 1. Timestamp: long denoting the time in ms; 2. All other fields are double. The following field will be generated for you automatically: Duration and Iteration.

Parameters:

scoreTable: - HashMap containing column headers and arraylist containing the history of values collected.

tableName: - title/name of your scoring table

startTime: - time your model building job was first started.

transformEigenVectors

public static double[][] transformEigenVectors(DataInfo dinfo,
                               double[][] vEigenIn)

This function will tranform the eigenvectors calculated for a matrix T(A) to the ones calculated for matrix A.

Parameters:

dinfo -

vEigenIn -

Returns: