[
https://issues.apache.org/jira/browse/MAHOUT-468?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12897924#action_12897924
]
Ted Dunning commented on MAHOUT-468:
------------------------------------
4) allocate many local variable in loop (need confirm )
In Class DistributedUncenteredZeroAssumingCosineVectorSimilarity
{code}
@Override
public double weight(Vector v) {
double length = 0.0;
Iterator<Element> elemIterator = v.iterateNonZero();
while (elemIterator.hasNext()) {
double value = elemIterator.next().get(); //this one
length += value * value; }
return Math.sqrt(length);
}
{code}
In fact the "allocation" of value here will have only a positive effect. In
general, primitives allocated in a small scope will be handled by the JIT in
about as optimal a fashion as is possible. Even heap-allocated structures in a
tight loop do not usually cause any measurable inefficiency (I have tested this
recently). In some cases, re-using a structure in a tight loop will be
*slower* than allocating new objects because it will cause the reused structure
to survive into the tenured generation while the multiple allocations will
simply cause a small amount of churn in the eden space.
> Performance of RowSimilarityJob is not good
> -------------------------------------------
>
> Key: MAHOUT-468
> URL: https://issues.apache.org/jira/browse/MAHOUT-468
> Project: Mahout
> Issue Type: Test
> Components: Collaborative Filtering
> Affects Versions: 0.4
> Reporter: Hui Wen Han
> Fix For: 0.4
>
>
> I have done a test ,
> Preferences records: 680,194
> distinct users: 23,246
> distinct items:437,569
> SIMILARITY_CLASS_NAME=SIMILARITY_COOCCURRENCE
> maybePruneItemUserMatrixPath:16.50M
> weights:13.80M
> pairwiseSimilarity:18.81G
> Job RowSimilarityJob-RowWeightMapper-WeightedOccurrencesPerColumnReducer:used
> 32 sec
> Job RowSimilarityJob-CooccurrencesMapper-SimilarityReducer:used 4.30 hours
> I think the reason may be following:
> 1) We used SequenceFileOutputFormat,it cause job can only be run by n ( n=
> Hadoop node counts ) mappers or reducers concurrently.
> 2) We stored redundant info.
> for example :
> the output of CooccurrencesMapper:
> (ItemIndexA,similarity),(ItemIndexA,ItemIndexB,similarity)
> 3) Some frequently used code
> https://issues.apache.org/jira/browse/MAHOUT-467
> 4) allocate many local variable in loop (need confirm )
> In Class DistributedUncenteredZeroAssumingCosineVectorSimilarity
> @Override
> public double weight(Vector v) {
> double length = 0.0;
> Iterator<Element> elemIterator = v.iterateNonZero();
> while (elemIterator.hasNext()) {
> double value = elemIterator.next().get(); //this one
> length += value * value;
> }
> return Math.sqrt(length);
> }
> 5) Maybe we need control the size of cooccurrences
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
