cascading.pipe.assembly
Class SumBy.SumPartials

java.lang.Object
  cascading.pipe.assembly.SumBy.SumPartials

All Implemented Interfaces:

AggregateBy.Functor, Serializable

Enclosing class:

SumBy

public static class SumBy.SumPartials
extends Object
implements AggregateBy.Functor

Class SumPartials is a AggregateBy.Functor that is used to sum observed duplicates from the tuple stream.

Use this class typically in tandem with a Sum Aggregator in order to improve counting performance by removing as many values as possible before the intermediate GroupBy operator.

See Also:

SumBy, Serialized Form

Constructor Summary

SumBy.SumPartials(Fields declaredFields)
Constructor SumPartials creates a new SumPartials instance.

SumBy.SumPartials(Fields declaredFields, Class sumType)

Method Summary

Tuple	aggregate(FlowProcess flowProcess, TupleEntry args, Tuple context) Method aggregate operates on the given args in tandem (optionally) with the given context values.
Tuple	complete(FlowProcess flowProcess, Tuple context) Method complete allows the final aggregate computation to be performed before the return value is collected.
Fields	getDeclaredFields() Method getDeclaredFields returns the declaredFields of this Functor object.

Methods inherited from class java.lang.Object

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

Constructor Detail

SumBy.SumPartials

public SumBy.SumPartials(Fields declaredFields)

Constructor SumPartials creates a new SumPartials instance.

SumBy.SumPartials

public SumBy.SumPartials(Fields declaredFields,
                         Class sumType)

Method Detail

getDeclaredFields

public Fields getDeclaredFields()

Description copied from interface: AggregateBy.Functor

Method getDeclaredFields returns the declaredFields of this Functor object.

Specified by:

getDeclaredFields in interface AggregateBy.Functor

Returns:

the declaredFields (type Fields) of this Functor object.

aggregate

public Tuple aggregate(FlowProcess flowProcess,
                       TupleEntry args,
                       Tuple context)

Description copied from interface: AggregateBy.Functor

Method aggregate operates on the given args in tandem (optionally) with the given context values.

The context argument is the result of the previous call to this method. Use it to store values between aggregate calls (the current count, or sum of the args).

On the very first invocation of aggregate for a given grouping key, context will be null. All subsequent invocations context will be the value returned on the previous invocation.

Specified by:

aggregate in interface AggregateBy.Functor

Parameters:

flowProcess - of type FlowProcess

args - of type TupleEntry

context - of type Tuple @return Tuple

complete

public Tuple complete(FlowProcess flowProcess,
                      Tuple context)

Description copied from interface: AggregateBy.Functor

Method complete allows the final aggregate computation to be performed before the return value is collected.

The number of values in the returned Tuple instance must match the number of declaredFields.

It is safe to return the context object as the result value.

Specified by:

complete in interface AggregateBy.Functor

Parameters:

flowProcess - of type FlowProcess

context - of type Tuple @return Tuple