Package org.apache.lucene.sandbox.facet.plain.histograms

Class HistogramCollectorManager

java.lang.Object

org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollectorManager

All Implemented Interfaces:

CollectorManager<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector,LongIntHashMap>

public final class HistogramCollectorManager
extends Object
implements CollectorManager<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector,LongIntHashMap>

CollectorManager that computes a histogram of the distribution of the values of a field.

It takes an bucketWidth as a parameter and counts the number of documents that fall into intervals [0, bucketWidth), [bucketWidth, 2*bucketWidth), etc. The keys of the returned LongIntHashMap identify these intervals as the quotient of the integer division by bucketWidth. Said otherwise, a key equal to k maps to values in the interval [k * bucketWidth, (k+1) * bucketWidth).

This implementation is optimized for the case when field is part of the index sort and has a skip index.

Note: this collector is inspired from "YU, Muzhi, LIN, Zhaoxiang, SUN, Jinan, et al. TencentCLS: the cloud log service with high query performances. Proceedings of the VLDB Endowment, 2022, vol. 15, no 12, p. 3472-3482.", where the authors describe how they run "histogram queries" by sorting the index by timestamp and pre-computing ranges of doc IDs for every possible bucket.

Constructor Summary

Constructors

Constructor	Description
HistogramCollectorManager(String field, long bucketWidth)	Compute a histogram of the distribution of the values of the given field according to the given bucketWidth.
HistogramCollectorManager(String field, long bucketWidth, int maxBuckets)	Expert constructor.

Method Summary

Modifier and Type	Method	Description
org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector	newCollector()
LongIntHashMap	reduce(Collection<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector> collectors)

Methods inherited from class java.lang.Object

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

Constructor Details

HistogramCollectorManager

public HistogramCollectorManager(String field,
 long bucketWidth)

Compute a histogram of the distribution of the values of the given field according to the given bucketWidth. This configures a maximum number of buckets equal to the default of 1024.

HistogramCollectorManager

public HistogramCollectorManager(String field,
 long bucketWidth,
 int maxBuckets)

Expert constructor.

Parameters:

maxBuckets - Max allowed number of buckets. Note that this is checked at runtime and on a best-effort basis.

Method Details

newCollector

public org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector newCollector()
                                                                                 throws IOException

Specified by:

newCollector in interface CollectorManager<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector,LongIntHashMap>

Throws:

IOException

reduce

public LongIntHashMap reduce(Collection<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector> collectors)
                      throws IOException

Specified by:

reduce in interface CollectorManager<org.apache.lucene.sandbox.facet.plain.histograms.HistogramCollector,LongIntHashMap>

Throws:

IOException