public class CPUAccumulationViaTensorTask extends BaseCPUTask<Double>
| Modifier and Type | Field and Description |
|---|---|
protected Accumulation |
op |
protected boolean |
outerTask |
protected List<Task<Double>> |
subTasks |
| Constructor and Description |
|---|
CPUAccumulationViaTensorTask(Accumulation op,
int threshold,
boolean outerTask) |
| Modifier and Type | Method and Description |
|---|---|
Double |
blockUntilComplete()
Assuming invokeAsync() has been called, block until the execution completes
|
Double |
call()
Compute the result immediately, in the current thread
|
protected Double |
compute() |
doTensorFirst, invokeAsync, invokeBlockingexec, getRawResult, setRawResultadapt, adapt, adapt, cancel, compareAndSetForkJoinTaskTag, complete, completeExceptionally, fork, get, get, getException, getForkJoinTaskTag, getPool, getQueuedTaskCount, getSurplusQueuedTaskCount, helpQuiesce, inForkJoinPool, invoke, invokeAll, invokeAll, invokeAll, isCancelled, isCompletedAbnormally, isCompletedNormally, isDone, join, peekNextLocalTask, pollNextLocalTask, pollTask, quietlyComplete, quietlyInvoke, quietlyJoin, reinitialize, setForkJoinTaskTag, tryUnforkprotected final Accumulation op
protected final boolean outerTask
public CPUAccumulationViaTensorTask(Accumulation op, int threshold, boolean outerTask)
public Double blockUntilComplete()
Taskpublic Double call()
Taskprotected Double compute()
compute in class RecursiveTask<Double>Copyright © 2016. All Rights Reserved.