hex.tree.xgboost

Class XGBoostUtils

java.lang.Object

hex.tree.xgboost.XGBoostUtils

public class XGBoostUtils
extends java.lang.Object

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	XGBoostUtils.FeatureProperties

Constructor Summary

Constructors

Constructor and Description
XGBoostUtils()

Method Summary

Modifier and Type	Method and Description
static XGBoostUtils.FeatureProperties	assembleFeatureNames(hex.DataInfo di)
static ai.h2o.xgboost4j.java.DMatrix	convertChunksToDMatrix(hex.DataInfo di, water.fvec.Chunk[] chunks, int response, boolean sparse, int offset) convert a set of H2O chunks (representing a part of a vector) to a sparse DMatrix
static MatrixLoader.DMatrixProvider	convertFrameToDMatrix(hex.DataInfo di, water.fvec.Frame frame, java.lang.String response, java.lang.String weight, java.lang.String offset, boolean sparse) convert an H2O Frame to a sparse DMatrix
static void	createFeatureMap(XGBoostModel model, water.fvec.Frame train)
static long	sumChunksLength(int[] chunkIds, water.fvec.Vec vec, java.util.Optional<water.fvec.Vec> weightsVector, int[] chunkLengths) Counts a total sum of chunks inside a vector.

Methods inherited from class java.lang.Object

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

Constructor Detail

XGBoostUtils

public XGBoostUtils()

Method Detail

createFeatureMap

public static void createFeatureMap(XGBoostModel model,
                                    water.fvec.Frame train)

convertFrameToDMatrix

public static MatrixLoader.DMatrixProvider convertFrameToDMatrix(hex.DataInfo di,
                                                                 water.fvec.Frame frame,
                                                                 java.lang.String response,
                                                                 java.lang.String weight,
                                                                 java.lang.String offset,
                                                                 boolean sparse)

convert an H2O Frame to a sparse DMatrix

Parameters:

di - data info

frame - H2O Frame - adapted using a provided data info

response - name of the response column

weight - name of the weight column

Returns:

DMatrix

Throws:

ai.h2o.xgboost4j.java.XGBoostError

sumChunksLength

public static long sumChunksLength(int[] chunkIds,
                                   water.fvec.Vec vec,
                                   java.util.Optional<water.fvec.Vec> weightsVector,
                                   int[] chunkLengths)

Counts a total sum of chunks inside a vector. Only chunks present in chunkIds are counted. If a weights vector is provided, only rows with non-zero weights are counted.

Parameters:

chunkIds - Chunk ids to consider during the calculation. Chunks IDs not listed are not included.

vec - Vector containing given chunk identifiers

weightsVector - Vector with row weights, possibly an empty optional

chunkLengths - Array of integers where the lengths of the individual chunks will be added. Initialization to an array of 0's is expected.

Returns:

A sum of chunk lengths. Possibly zero, if there are no chunks or the chunks are empty.

convertChunksToDMatrix

public static ai.h2o.xgboost4j.java.DMatrix convertChunksToDMatrix(hex.DataInfo di,
                                                                   water.fvec.Chunk[] chunks,
                                                                   int response,
                                                                   boolean sparse,
                                                                   int offset)
                                                            throws ai.h2o.xgboost4j.java.XGBoostError

convert a set of H2O chunks (representing a part of a vector) to a sparse DMatrix

Returns:

DMatrix

Throws:

ai.h2o.xgboost4j.java.XGBoostError

assembleFeatureNames

public static XGBoostUtils.FeatureProperties assembleFeatureNames(hex.DataInfo di)