public class DecisionTreeMetadata
extends Object
implements scala.Serializable
Constructor and Description |
---|
DecisionTreeMetadata(int numFeatures,
long numExamples,
int numClasses,
int maxBins,
scala.collection.immutable.Map<Object,Object> featureArity,
scala.collection.immutable.Set<Object> unorderedFeatures,
int[] numBins,
Impurity impurity,
scala.Enumeration.Value quantileStrategy,
int maxDepth,
int minInstancesPerNode,
double minInfoGain,
int numTrees,
int numFeaturesPerNode) |
Modifier and Type | Method and Description |
---|---|
static DecisionTreeMetadata |
buildMetadata(RDD<LabeledPoint> input,
Strategy strategy)
|
static DecisionTreeMetadata |
buildMetadata(RDD<LabeledPoint> input,
Strategy strategy,
int numTrees,
String featureSubsetStrategy)
Construct a
DecisionTreeMetadata instance for this dataset and parameters. |
scala.collection.immutable.Map<Object,Object> |
featureArity() |
Impurity |
impurity() |
boolean |
isCategorical(int featureIndex) |
boolean |
isClassification() |
boolean |
isContinuous(int featureIndex) |
boolean |
isMulticlass() |
boolean |
isMulticlassWithCategoricalFeatures() |
boolean |
isUnordered(int featureIndex) |
int |
maxBins() |
int |
maxDepth() |
double |
minInfoGain() |
int |
minInstancesPerNode() |
int[] |
numBins() |
int |
numClasses() |
long |
numExamples() |
int |
numFeatures() |
int |
numFeaturesPerNode() |
int |
numSplits(int featureIndex)
Number of splits for the given feature.
|
int |
numTrees() |
static int |
numUnorderedBins(int arity)
Given the arity of a categorical feature (arity = number of categories),
return the number of bins for the feature if it is to be treated as an unordered feature.
|
scala.Enumeration.Value |
quantileStrategy() |
void |
setNumSplits(int featureIndex,
int numSplits)
Set number of splits for a continuous feature.
|
boolean |
subsamplingFeatures()
Indicates if feature subsampling is being used.
|
scala.collection.immutable.Set<Object> |
unorderedFeatures() |
public DecisionTreeMetadata(int numFeatures, long numExamples, int numClasses, int maxBins, scala.collection.immutable.Map<Object,Object> featureArity, scala.collection.immutable.Set<Object> unorderedFeatures, int[] numBins, Impurity impurity, scala.Enumeration.Value quantileStrategy, int maxDepth, int minInstancesPerNode, double minInfoGain, int numTrees, int numFeaturesPerNode)
public static DecisionTreeMetadata buildMetadata(RDD<LabeledPoint> input, Strategy strategy, int numTrees, String featureSubsetStrategy)
DecisionTreeMetadata
instance for this dataset and parameters.
This computes which categorical features will be ordered vs. unordered,
as well as the number of splits and bins for each feature.public static DecisionTreeMetadata buildMetadata(RDD<LabeledPoint> input, Strategy strategy)
public static int numUnorderedBins(int arity)
public int numFeatures()
public long numExamples()
public int numClasses()
public int maxBins()
public scala.collection.immutable.Map<Object,Object> featureArity()
public scala.collection.immutable.Set<Object> unorderedFeatures()
public int[] numBins()
public Impurity impurity()
public scala.Enumeration.Value quantileStrategy()
public int maxDepth()
public int minInstancesPerNode()
public double minInfoGain()
public int numTrees()
public int numFeaturesPerNode()
public boolean isUnordered(int featureIndex)
public boolean isClassification()
public boolean isMulticlass()
public boolean isMulticlassWithCategoricalFeatures()
public boolean isCategorical(int featureIndex)
public boolean isContinuous(int featureIndex)
public int numSplits(int featureIndex)
public void setNumSplits(int featureIndex, int numSplits)
public boolean subsamplingFeatures()