org.apache.cassandra.utils

Class BloomCalculations

java.lang.Object

org.apache.cassandra.utils.BloomCalculations

public class BloomCalculations
extends java.lang.Object

The following calculations are taken from: http://www.cs.wisc.edu/~cao/papers/summary-cache/node8.html "Bloom Filters - the math" This class's static methods are meant to facilitate the use of the Bloom Filter class by helping to choose correct values of 'bits per element' and 'number of hash functions, k'.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	BloomCalculations.BloomSpecification A wrapper class that holds two key parameters for a Bloom Filter: the number of hash functions used, and the number of buckets per element used.

Constructor Summary

Constructors

Constructor and Description
BloomCalculations()

Method Summary

Methods

Modifier and Type	Method and Description
static BloomCalculations.BloomSpecification	computeBloomSpec(int bucketsPerElement) Given the number of buckets that can be used per element, return a specification that minimizes the false positive rate.
static BloomCalculations.BloomSpecification	computeBloomSpec(int maxBucketsPerElement, double maxFalsePosProb) Given a maximum tolerable false positive probability, compute a Bloom specification which will give less than the specified false positive rate, but minimize the number of buckets per element and the number of hash functions used.
static int	maxBucketsPerElement(long numElements) Calculates the maximum number of buckets per element that this implementation can support.
static double	minSupportedBloomFilterFpChance() Retrieves the minimum supported BloomFilterFpChance value

Methods inherited from class java.lang.Object

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

Constructor Detail

BloomCalculations

public BloomCalculations()

Method Detail

computeBloomSpec

public static BloomCalculations.BloomSpecification computeBloomSpec(int bucketsPerElement)

Given the number of buckets that can be used per element, return a specification that minimizes the false positive rate.

Parameters:

bucketsPerElement - The number of buckets per element for the filter.

Returns:

A spec that minimizes the false positive rate.

computeBloomSpec

public static BloomCalculations.BloomSpecification computeBloomSpec(int maxBucketsPerElement,
                                                    double maxFalsePosProb)

Given a maximum tolerable false positive probability, compute a Bloom specification which will give less than the specified false positive rate, but minimize the number of buckets per element and the number of hash functions used. Because bandwidth (and therefore total bitvector size) is considered more expensive than computing power, preference is given to minimizing buckets per element rather than number of hash functions.

Parameters:

maxBucketsPerElement - The maximum number of buckets available for the filter.

maxFalsePosProb - The maximum tolerable false positive rate.

Returns:

A Bloom Specification which would result in a false positive rate less than specified by the function call

Throws:

java.lang.UnsupportedOperationException - if a filter satisfying the parameters cannot be met

maxBucketsPerElement

public static int maxBucketsPerElement(long numElements)

Calculates the maximum number of buckets per element that this implementation can support. Crucially, it will lower the bucket count if necessary to meet BitSet's size restrictions.

minSupportedBloomFilterFpChance

public static double minSupportedBloomFilterFpChance()

Retrieves the minimum supported BloomFilterFpChance value

Returns:

Minimum supported value for BloomFilterFpChance