hex

Class FrameTask.DataInfo

java.lang.Object

water.Iced

hex.FrameTask.DataInfo

All Implemented Interfaces:

java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, water.Freezable

Enclosing class:

FrameTask<T extends FrameTask<T>>

public static class FrameTask.DataInfo
extends water.Iced

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	FrameTask.DataInfo.TransformType

Field Summary

Fields

Modifier and Type	Field and Description
water.fvec.Frame	_adaptedFrame
int[]	_catMissing
int[]	_catOffsets
int	_cats
int	_foldId
water.Key	_frameKey
int	_nfolds
double[]	_normMul
double[]	_normRespMul
double[]	_normRespSub
double[]	_normSub
int	_nums
FrameTask.DataInfo.TransformType	_predictor_transform
FrameTask.DataInfo.TransformType	_response_transform
int	_responses
boolean	_useAllFactorLevels

Constructor Summary

Constructors

Constructor and Description
FrameTask.DataInfo(water.fvec.Frame fr, int[][] catLevels, int responses, FrameTask.DataInfo.TransformType predictor_transform, FrameTask.DataInfo.TransformType response_transform, int foldId, int nfolds)
FrameTask.DataInfo(water.fvec.Frame fr, int hasResponses, boolean useAllFactorLvls, double[] normSub, double[] normMul, FrameTask.DataInfo.TransformType predictor_transform, double[] normRespSub, double[] normRespMul)
FrameTask.DataInfo(water.fvec.Frame fr, int nResponses, boolean useAllFactors, FrameTask.DataInfo.TransformType predictor_transform)
FrameTask.DataInfo(water.fvec.Frame fr, int nResponses, boolean useAllFactorLevels, FrameTask.DataInfo.TransformType predictor_transform, FrameTask.DataInfo.TransformType response_transform)

Method Summary

Methods

Modifier and Type	Method and Description
java.lang.String[]	coefNames()
FrameTask.DataInfo	deep_clone()
FrameTask.DataInfo	filterExpandedColumns(int[] cols)
int	fullN()
FrameTask.DataInfo	getFold(int foldId, int nfolds)
int	largestCat()
int	numStart()
static water.fvec.Frame	prepareFrame(water.fvec.Frame source, int[] ignored_cols, boolean dropConstantCols, boolean dropNACols)
static water.fvec.Frame	prepareFrame(water.fvec.Frame source, water.fvec.Vec response, int[] ignored_cols, boolean toEnum, boolean dropConstantCols)
static water.fvec.Frame	prepareFrame(water.fvec.Frame source, water.fvec.Vec response, int[] ignored_cols, boolean toEnum, boolean dropConstantCols, boolean dropNACols) Prepare a Frame (with a single response) to be processed by the FrameTask 1) Place response at the end 2) (Optionally) Remove columns with constant values or with greater than 20% NaNs 3) Possibly turn integer categoricals into enums
void	softMaxCategoricals(float[] in, float[] out) Normalize horizontalized categoricals to become probabilities per factor level.
java.lang.String	toString()
void	unScaleNumericals(float[] in, float[] out) Undo the standardization/normalization of numerical columns

Methods inherited from class water.Iced

clone, frozenType, read_impl, read, readExternal, readJSON_impl, readJSON, write_impl, write, writeExternal, writeHTML_impl, writeHTML, writeJSON_impl, writeJSON

Methods inherited from class java.lang.Object

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

Field Detail

_adaptedFrame

public water.fvec.Frame _adaptedFrame

_responses

public int _responses

_predictor_transform

public FrameTask.DataInfo.TransformType _predictor_transform

_response_transform

public FrameTask.DataInfo.TransformType _response_transform

_useAllFactorLevels

public boolean _useAllFactorLevels

_nums

public int _nums

_cats

public int _cats

_catOffsets

public int[] _catOffsets

_catMissing

public int[] _catMissing

_normMul

public double[] _normMul

_normSub

public double[] _normSub

_normRespMul

public double[] _normRespMul

_normRespSub

public double[] _normRespSub

_foldId

public int _foldId

_nfolds

public int _nfolds

_frameKey

public water.Key _frameKey

Constructor Detail

FrameTask.DataInfo

public FrameTask.DataInfo(water.fvec.Frame fr,
                  int hasResponses,
                  boolean useAllFactorLvls,
                  double[] normSub,
                  double[] normMul,
                  FrameTask.DataInfo.TransformType predictor_transform,
                  double[] normRespSub,
                  double[] normRespMul)

FrameTask.DataInfo

public FrameTask.DataInfo(water.fvec.Frame fr,
                  int nResponses,
                  boolean useAllFactors,
                  FrameTask.DataInfo.TransformType predictor_transform)

FrameTask.DataInfo

public FrameTask.DataInfo(water.fvec.Frame fr,
                  int[][] catLevels,
                  int responses,
                  FrameTask.DataInfo.TransformType predictor_transform,
                  FrameTask.DataInfo.TransformType response_transform,
                  int foldId,
                  int nfolds)

FrameTask.DataInfo

public FrameTask.DataInfo(water.fvec.Frame fr,
                  int nResponses,
                  boolean useAllFactorLevels,
                  FrameTask.DataInfo.TransformType predictor_transform,
                  FrameTask.DataInfo.TransformType response_transform)

Method Detail

deep_clone

public FrameTask.DataInfo deep_clone()

prepareFrame

public static water.fvec.Frame prepareFrame(water.fvec.Frame source,
                            water.fvec.Vec response,
                            int[] ignored_cols,
                            boolean toEnum,
                            boolean dropConstantCols,
                            boolean dropNACols)

Prepare a Frame (with a single response) to be processed by the FrameTask 1) Place response at the end 2) (Optionally) Remove columns with constant values or with greater than 20% NaNs 3) Possibly turn integer categoricals into enums

Parameters:

source - A frame to be expanded and sanity checked

response - (should be part of source)

toEnum - Whether or not to turn categoricals into enums

dropConstantCols - Whether or not to drop constant columns

Returns:

Frame to be used by FrameTask

prepareFrame

public static water.fvec.Frame prepareFrame(water.fvec.Frame source,
                            int[] ignored_cols,
                            boolean dropConstantCols,
                            boolean dropNACols)

prepareFrame

public static water.fvec.Frame prepareFrame(water.fvec.Frame source,
                            water.fvec.Vec response,
                            int[] ignored_cols,
                            boolean toEnum,
                            boolean dropConstantCols)

filterExpandedColumns

public FrameTask.DataInfo filterExpandedColumns(int[] cols)

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

getFold

public FrameTask.DataInfo getFold(int foldId,
                         int nfolds)

fullN

public final int fullN()

largestCat

public final int largestCat()

numStart

public final int numStart()

coefNames

public final java.lang.String[] coefNames()

softMaxCategoricals

public final void softMaxCategoricals(float[] in,
                       float[] out)

Normalize horizontalized categoricals to become probabilities per factor level. This is done with the SoftMax function.

Parameters:

in - input values

out - output values (can be the same as input)

unScaleNumericals

public final void unScaleNumericals(float[] in,
                     float[] out)

Undo the standardization/normalization of numerical columns

Parameters:

in - input values

out - output values (can be the same as input)