edu.iu.dsc.tws.tset.links.batch

Class BatchGatherLink<T>

java.lang.Object

edu.iu.dsc.tws.tset.links.BaseTLink<T1,T0>

edu.iu.dsc.tws.tset.links.BaseTLinkWithSchema<T1,T0>

edu.iu.dsc.tws.tset.links.batch.BatchTLinkImpl<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>

edu.iu.dsc.tws.tset.links.batch.BatchGatherLink<T>

Type Parameters:

T - value type

All Implemented Interfaces:

BatchTLink<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>, TLink<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>, StoringData<T>, TBase, Buildable, BuildableTLink, java.io.Serializable

Direct Known Subclasses:

AllGatherTLink, GatherTLink

public abstract class BatchGatherLink<T>
extends BatchTLinkImpl<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>

This is the Tlinks used by gather operations. Specific operations such as map, flatmap, cache, etc will be done on the tuple value only (key will be dropped, as key is an information forcibly attached at the communication level). If the key information is required, users can use the compute methods which enables the use of Iterator<Tuple<Integer, T>>

See Also:

Serialized Form

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	BatchGatherLink()

Method Summary

Modifier and Type	Method and Description
CachedTSet<T>	cache() Runs this TSet and caches the data to an in-memory DataPartition and exposes the data as another TSet.
<O> ComputeTSet<O,java.util.Iterator<Tuple<java.lang.Integer,T>>>	flatmap(FlatMapFunc<O,T> mapFn) Performs flat map operation based on the FlatMapFunc provided
void	forEach(ApplyFunc<T> applyFunction) Applies a function elementwise.
CachedTSet<T>	lazyCache() Performs caching lazily.
ComputeTSet<java.lang.Object,java.util.Iterator<Tuple<java.lang.Integer,T>>>	lazyForEach(ApplyFunc<T> applyFunction)
PersistedTSet<T>	lazyPersist() Performs persisting lazily.
<O> ComputeTSet<O,java.util.Iterator<Tuple<java.lang.Integer,T>>>	map(MapFunc<O,T> mapFn) Performs elementwise map operation based on the MapFunc provided
<K,V> KeyedTSet<K,V>	mapToTuple(MapFunc<Tuple<K,V>,T> genTupleFn) Maps the data passed through the TLink to Tuple based on a MapFunc and thereby creating a TupleTSet
PersistedTSet<T>	persist() Similar to cache, but the data is stored in a disk based DataPartition.

Methods inherited from class edu.iu.dsc.tws.tset.links.batch.BatchTLinkImpl

compute, compute, compute, compute, getTSetEnv, sink

Methods inherited from class edu.iu.dsc.tws.tset.links.BaseTLinkWithSchema

getSchema

Methods inherited from class edu.iu.dsc.tws.tset.links.BaseTLink

addChildToGraph, equals, getId, getName, getSourceParallelism, getTargetParallelism, hashCode, rename, setTSetEnv, toString

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Methods inherited from interface edu.iu.dsc.tws.api.tset.link.batch.BatchTLink

setName

Methods inherited from interface edu.iu.dsc.tws.api.tset.TBase

getId, getName

Methods inherited from interface edu.iu.dsc.tws.tset.links.BuildableTLink

build, getEdge

Methods inherited from interface edu.iu.dsc.tws.tset.Buildable

generateID, getTBaseGraph

Constructor Detail

BatchGatherLink

protected BatchGatherLink()

Method Detail

map

public <O> ComputeTSet<O,java.util.Iterator<Tuple<java.lang.Integer,T>>> map(MapFunc<O,T> mapFn)

Description copied from interface: TLink

Performs elementwise map operation based on the MapFunc provided

Type Parameters:

O - output type

Parameters:

mapFn - map function T0 to O

Returns:

Compute tset

flatmap

public <O> ComputeTSet<O,java.util.Iterator<Tuple<java.lang.Integer,T>>> flatmap(FlatMapFunc<O,T> mapFn)

Description copied from interface: TLink

Performs flat map operation based on the FlatMapFunc provided

Type Parameters:

O - map function to T0 to multiple elements of <O>

Parameters:

mapFn - map function which can produce multiple elements for a single <T0> element

Returns:

Compute TSet

mapToTuple

public <K,V> KeyedTSet<K,V> mapToTuple(MapFunc<Tuple<K,V>,T> genTupleFn)

Description copied from interface: TLink

Maps the data passed through the TLink to Tuple based on a MapFunc and thereby creating a TupleTSet

Type Parameters:

K - key type

V - value type

Parameters:

genTupleFn - MapFunc to generate a Tuple

Returns:

Keyed TSet

forEach

public void forEach(ApplyFunc<T> applyFunction)

Description copied from interface: TLink

Applies a function elementwise. Similar to compute, but the ApplyFunc does not return anything.

Parameters:

applyFunction - apply function

lazyForEach

public ComputeTSet<java.lang.Object,java.util.Iterator<Tuple<java.lang.Integer,T>>> lazyForEach(ApplyFunc<T> applyFunction)

lazyCache

public CachedTSet<T> lazyCache()

Description copied from interface: StoringData

Performs caching lazily. i.e. cache operation would only be performed when the TSet is evaluated explicitly.

Returns:

Cached TSet

cache

public CachedTSet<T> cache()

Description copied from interface: StoringData

Runs this TSet and caches the data to an in-memory DataPartition and exposes the data as another TSet.

Specified by:

cache in interface StoringData<T>

Overrides:

cache in class BatchTLinkImpl<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>

Returns:

Cached TSet

lazyPersist

public PersistedTSet<T> lazyPersist()

Description copied from interface: StoringData

Performs persisting lazily.

Returns:

Persisted / Checkpointed TSets

persist

public PersistedTSet<T> persist()

Description copied from interface: StoringData

Similar to cache, but the data is stored in a disk based DataPartition. This method would also expose the checkpointing ability to TSets.

Specified by:

persist in interface StoringData<T>

Overrides:

persist in class BatchTLinkImpl<java.util.Iterator<Tuple<java.lang.Integer,T>>,T>

Returns:

Persisted / Checkpointed TSets