org.datavec.api.transform.ops

Class AggregatorImpls.AggregableCountUnique<T>

java.lang.Object

org.datavec.api.transform.ops.AggregatorImpls.AggregableCountUnique<T>

Type Parameters:

T -

All Implemented Interfaces:

Serializable, IAggregableReduceOp<T,Writable>, org.nd4j.linalg.function.Consumer<T>, org.nd4j.linalg.function.Supplier<Writable>

Enclosing class:

AggregatorImpls

public static class AggregatorImpls.AggregableCountUnique<T>
extends Object
implements IAggregableReduceOp<T,Writable>

This distinct count is based on streamlib's implementation of "HyperLogLog in Practice: Algorithmic Engineering of a State of The Art Cardinality Estimation Algorithm", available here. The relative accuracy is approximately `1.054 / sqrt(2^p)`. Setting a nonzero `sp > p` in HyperLogLogPlus(p, sp) would trigger sparse representation of registers, which may reduce the memory consumption and increase accuracy when the cardinality is small.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
AggregableCountUnique(float precision)

Method Summary

Modifier and Type	Method and Description
void	accept(T element)
<U extends IAggregableReduceOp<T,Writable>> void	combine(U acc)
Writable	get()

Methods inherited from class java.lang.Object

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

Constructor Detail

AggregableCountUnique

public AggregableCountUnique(float precision)

Method Detail

accept

public void accept(T element)

Specified by:

accept in interface org.nd4j.linalg.function.Consumer<T>

combine

public <U extends IAggregableReduceOp<T,Writable>> void combine(U acc)

Specified by:

combine in interface IAggregableReduceOp<T,Writable>

get

public Writable get()

Specified by:

get in interface org.nd4j.linalg.function.Supplier<Writable>