Browse Benchmarks
Benchmark ::implicit_unary_grouping : Corelated nested query with aggregate function in the return clause.
The correlation predicate performs general comparison on one sequence-valued
argument.
(submitted at 2006-11-24) |
|---|
| Authors: |
Norman May (University of Mannheim) |
| Categories: |
XQuery,
|
| Target: |
All systems |
| Engine type: |
persistent-storage main-memory |
| Measure: |
The measure targets the total execution time.
|
Result unit: |
Milliseconds |
| Query: |
|---|
|
The naive evaluation of this query evaluates in nested-loops.
This results in a quadratic increase in excution time when the document size is increased.
More efficient evaluation strategies scale better.
In particular, this query can be evaluated with unary grouping.
|
|
Syntax: |
let $d1 := doc("bib.xml")
for $a1 in distinct-values($d1//author)
return
<author>
<name> { $a1 } </name>
{
for $b2 in $d1/book[$a1 = author]
return $b2/title
}
</author>
|
|
Language: |
XQuery 1.0 |
| Document: |
|---|
| The document creates a document that conforms to the DTD used for the XQuery Use Case XMP. |
|
Generator: |
Toxgene and xmp.tsl |
| Parameters: |
|---|
Parameter
num_books:
The number of books, authors, reviews, prices.
This is the main scaling parameter
|
|
This parameter characterizes the: |
doc |
|
Unit: |
None. |
|
Values: |
100 1000 10000 |
|
Distribution: |
uniform |
|
Scale: |
ordinal |
|
|
|
Parameter
num_authors_per_book:
The number of authors per book. |
|
This parameter characterizes the: |
doc |
|
Unit: |
None |
|
Values: |
2 5 10 |
|
Distribution: |
uniform |
|
Scale: |
ordinal |
|
|
|
|
| Methodology: |
|
Running scenario: queryScale
|
|---|
|
Test the influence of the document size (which mainly depends on
parameter num_books) on the query execution time.
|
| Parameter instantiation: |
|
Name |
Value(s) |
|
num_books |
100
,
1000
,
10000 |
|
|
Download the micro-benchmark:
[XML]
|
|