org.apache.beam.sdk.extensions.zetasketch

Class HllCount.Init

java.lang.Object

org.apache.beam.sdk.extensions.zetasketch.HllCount.Init

Enclosing class:

HllCount

public static final class HllCount.Init
extends java.lang.Object

Provides PTransforms to aggregate inputs into HLL++ sketches. The four supported input types are Integer, Long, String, and byte[].

Sketches are represented using the byte[] type. Sketches of the same type can be merged into a new sketch using HllCount.MergePartial. Estimated count of distinct elements can be extracted from sketches using HllCount.Extract.

An "empty sketch" represented by an byte array of length 0 is returned if the input PCollection is empty.

Corresponds to the HLL_COUNT.INIT(input [, precision]) function in BigQuery.

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	HllCount.Init.Builder<InputT> Builder for the HllCount.Init combining PTransform.

Method Summary

Modifier and Type	Method and Description
static HllCount.Init.Builder<byte[]>	forBytes() Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes bytes-type HLL++ sketches.
static HllCount.Init.Builder<java.lang.Integer>	forIntegers() Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes integer-type HLL++ sketches.
static HllCount.Init.Builder<java.lang.Long>	forLongs() Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes long-type HLL++ sketches.
static HllCount.Init.Builder<java.lang.String>	forStrings() Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes string-type HLL++ sketches.

Methods inherited from class java.lang.Object

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

Method Detail

forIntegers

public static HllCount.Init.Builder<java.lang.Integer> forIntegers()

Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes integer-type HLL++ sketches. Call HllCount.Init.Builder.globally() or HllCount.Init.Builder.perKey() on the returning HllCount.Init.Builder to finalize the PTransform.

Calling HllCount.Init.Builder.globally() returns a PTransform that takes an input PCollection<Integer> and returns a PCollection<byte[]> which consists of the integer-type HLL++ sketch computed from the elements in the input PCollection.

Calling HllCount.Init.Builder.perKey() returns a PTransform that takes an input PCollection<KV<K, Integer>> and returns a PCollection<KV<K, byte[]>> which consists of the per-key integer-type HLL++ sketch computed from the values matching each key in the input PCollection.

Integer-type sketches cannot be merged with sketches of other types.

forLongs

public static HllCount.Init.Builder<java.lang.Long> forLongs()

Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes long-type HLL++ sketches. Call HllCount.Init.Builder.globally() or HllCount.Init.Builder.perKey() on the returning HllCount.Init.Builder to finalize the PTransform.

Calling HllCount.Init.Builder.globally() returns a PTransform that takes an input PCollection<Long> and returns a PCollection<byte[]> which consists of the long-type HLL++ sketch computed from the elements in the input PCollection.

Calling HllCount.Init.Builder.perKey() returns a PTransform that takes an input PCollection<KV<K, Long>> and returns a PCollection<KV<K, byte[]>> which consists of the per-key long-type HLL++ sketch computed from the values matching each key in the input PCollection.

Long-type sketches cannot be merged with sketches of other types.

forStrings

public static HllCount.Init.Builder<java.lang.String> forStrings()

Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes string-type HLL++ sketches. Call HllCount.Init.Builder.globally() or HllCount.Init.Builder.perKey() on the returning HllCount.Init.Builder to finalize the PTransform.

Calling HllCount.Init.Builder.globally() returns a PTransform that takes an input PCollection<String> and returns a PCollection<byte[]> which consists of the string-type HLL++ sketch computed from the elements in the input PCollection.

Calling HllCount.Init.Builder.perKey() returns a PTransform that takes an input PCollection<KV<K, String>> and returns a PCollection<KV<K, byte[]>> which consists of the per-key string-type HLL++ sketch computed from the values matching each key in the input PCollection.

String-type sketches cannot be merged with sketches of other types.

forBytes

public static HllCount.Init.Builder<byte[]> forBytes()

Returns a HllCount.Init.Builder for a HllCount.Init combining PTransform that computes bytes-type HLL++ sketches. Call HllCount.Init.Builder.globally() or HllCount.Init.Builder.perKey() on the returning HllCount.Init.Builder to finalize the PTransform.

Calling HllCount.Init.Builder.globally() returns a PTransform that takes an input PCollection<byte[]> and returns a PCollection<byte[]> which consists of the bytes-type HLL++ sketch computed from the elements in the input PCollection.

Calling HllCount.Init.Builder.perKey() returns a PTransform that takes an input PCollection<KV<K, byte[]>> and returns a PCollection<KV<K, byte[]>> which consists of the per-key bytes-type HLL++ sketch computed from the values matching each key in the input PCollection.

Bytes-type sketches cannot be merged with sketches of other types.