org.apache.solr.ltr.feature

Class Feature.FeatureWeight

java.lang.Object

org.apache.lucene.search.Weight

org.apache.solr.ltr.feature.Feature.FeatureWeight

Direct Known Subclasses:

FieldLengthFeature.FieldLengthFeatureWeight, FieldValueFeature.FieldValueFeatureWeight, OriginalScoreFeature.OriginalScoreWeight, SolrFeature.SolrFeatureWeight, ValueFeature.ValueFeatureWeight

Enclosing class:

Feature

public abstract class Feature.FeatureWeight
extends Weight

Weight for a feature

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	Feature.FeatureWeight.FeatureScorer A 'recipe' for computing a feature
class	Feature.FeatureWeight.ValueFeatureScorer Default FeatureScorer class that returns the score passed in.

Nested classes/interfaces inherited from class org.apache.lucene.search.Weight

Weight.DefaultBulkScorer

Field Summary

Fields

Modifier and Type	Field and Description
protected Map<String,String[]>	efi
protected MacroExpander	macroExpander
protected Query	originalQuery
protected SolrQueryRequest	request
protected IndexSearcher	searcher

Fields inherited from class org.apache.lucene.search.Weight

parentQuery

Constructor Summary

Constructors

Constructor and Description
FeatureWeight(Query q, IndexSearcher searcher, SolrQueryRequest request, Query originalQuery, Map<String,String[]> efi) Initialize a feature without the normalizer from the feature file.

Method Summary

Modifier and Type	Method and Description
Explanation	explain(LeafReaderContext context, int doc)
void	extractTerms(Set<Term> terms)
float	getDefaultValue()
int	getIndex()
String	getName()
abstract Feature.FeatureWeight.FeatureScorer	scorer(LeafReaderContext context)
String	toString() Used in the FeatureWeight's explain.

Methods inherited from class org.apache.lucene.search.Weight

bulkScorer, getQuery, scorerSupplier

Methods inherited from class java.lang.Object

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

Field Detail

searcher

protected final IndexSearcher searcher

request

protected final SolrQueryRequest request

efi

protected final Map<String,String[]> efi

macroExpander

protected final MacroExpander macroExpander

originalQuery

protected final Query originalQuery

Constructor Detail

FeatureWeight

public FeatureWeight(Query q,
                     IndexSearcher searcher,
                     SolrQueryRequest request,
                     Query originalQuery,
                     Map<String,String[]> efi)

Initialize a feature without the normalizer from the feature file. This is called on initial construction since multiple models share the same features, but have different normalizers. A concrete model's feature is copied through featForNewModel().

Parameters:

q - Solr query associated with this FeatureWeight

searcher - Solr searcher available for features if they need them

Method Detail

getName

public String getName()

getIndex

public int getIndex()

getDefaultValue

public float getDefaultValue()

scorer

public abstract Feature.FeatureWeight.FeatureScorer scorer(LeafReaderContext context)
                                                    throws IOException

Specified by:

scorer in class Weight

Throws:

IOException

explain

public Explanation explain(LeafReaderContext context,
                           int doc)
                    throws IOException

Specified by:

explain in class Weight

Throws:

IOException

toString

public String toString()

Used in the FeatureWeight's explain. Each feature should implement this returning properties of the specific scorer useful for an explain. For example "MyCustomClassFeature [name=" + name + "myVariable:" + myVariable + "]"; If not provided, a default implementation will return basic feature properties, which might not include query time specific values.

Overrides:

toString in class Object

extractTerms

public void extractTerms(Set<Term> terms)

Specified by:

extractTerms in class Weight