Skip Headers
Oracle® CEP CQL Language Reference
11g Release 1 (11.1.1)
Part Number E12048-01
Go to Documentation Home
Home		 Go to Book List
Book List		 Go to Table of Contents
Contents		 Go to Index
Index		 Go to Feedback page
Contact Us
Go to previous page
Previous		 Go to next page
Next
View PDF

6 Functions: Aggregate

Aggregate functions perform a summary operation on all the values that a query returns.

For more information, see Section 1.1.9, "Functions".

6.1 Introduction to Oracle CQL Built-In Aggregate Functions

Table 6-1 lists the built-in aggregate functions that Oracle CQL provides:

Table 6-1 Oracle CQL Built-in Aggregate Functions

Type Function

Aggregate

Aggregate (incremental computation)

Extended aggregate

builtin_aggr::=

Surrounding text describes builtin_aggr.gif.

builtin_aggr_incr::=

Surrounding text describes builtin_aggr_incr.gif.

extended_builtin_aggr::=

Surrounding text describes extended_builtin_aggr.gif.

Specify distinct if you want Oracle CEP to return only one copy of each set of duplicate tuples selected. Duplicate tuples are those with matching values for each expression in the select list. For more information, see aggr_distinct_expr.

Oracle CEP does not support nested aggregations.

Note:

In stream input examples, lines beginning with h (such as h 3800) are heartbeat input tuples. These inform Oracle CEP that no further input will have a timestamp lesser than the heartbeat value.

For more information, see:

AVG

Syntax

Surrounding text describes avg.gif.

Purpose

AVG returns average value of expr.

This function takes as an argument any bigint, float, or int datatype. The function returns a float regardless of the numeric datatype of the argument.

Examples

Consider the query float_avg in Example 6-1 and the data stream S3 in Example 6-2. Stream S3 has schema (c1 float). The query returns the relation in Example 6-3. Note that the AVG function returns a result of NaN if the average value is not a number. For more information, see Section 2.4.2, "Numeric Literals".

Example 6-1 AVG Function Query

<query id="float_avg"><![CDATA[ 
    select avg(c1) from S3[range 5]
]]></query>

Example 6-2 AVG Function Stream Input

Timestamp   Tuple
 1000 
 2000        5.5
 8000        4.4
 9000
15000       44.2
h 200000000

Example 6-3 AVG Function Relation Output

Timestamp   Tuple Kind  Tuple
    0:      + 
 1000:      - 
 1000:      +            0.0
 2000:      -            0.0
 2000:      +            5.5
 6000:      -            5.5
 6000:      +            5.5
 7000:      -            5.5
 7000:      +
 8000:      -
 8000:      +            4.4
 9000:      -            4.4
 9000:      +            4.4
13000:      -            4.4
13000:      +           NaN
14000:      -           NaN
14000:      +
15000:      -
15000:      +           44.2
20000:      -           44.2
20000:      +

COUNT

Syntax

Surrounding text describes count.gif.

Purpose

COUNT returns the number of tuples returned by the query as an int value.

If you specify expr, then COUNT returns the number of tuples where expr is not null.

If you specify the asterisk (*), then this function returns a count of all tuples, including duplicates and nulls. COUNT never returns null.

Example

Consider the query q2 in Example 6-4 and the data stream S2 in Example 6-5. Stream S2 has schema (c1 integer, c2 integer). The query returns the relation in Example 6-6. For more information on range windows, see "Range-Based Stream-to-Relation Window Operators".

Example 6-4 COUNT Function Query

<query id="q2"><![CDATA[ 
    SELECT COUNT(c2), COUNT(*) FROM  S [RANGE 10]
]]></query>

Example 6-5 COUNT Function Stream Input

Timestamp   Tuple
1000        1,2
2000        1,
3000        1,4
6000        1,6

Example 6-6 COUNT Function Relation Output

Timestamp   Tuple Kind  Tuple
0:          +           0,0
1000:       -           0,0
1000:       +           1,1
2000:       +           1,1
2000:       +           2,2
3000:       +           2,2
3000:       +           3,3
5000:       +           3,3
5000:       +           3,4
6000:       +           3,4
6000:       +           4,5
7000:       -           4,5
7000:       +           5,6
8000:       +           5,6

FIRST

Syntax

Surrounding text describes first.gif.

Purpose

FIRST returns the value of the specified stream element the first time the specified pattern is matched.

The type of the specified stream element may be any of:

The return type of this function depends on the type of the specified stream element.

This function takes a single argument made up of the following period-separated values:

See Also:

Examples

Consider the query q9 in Example 6-7 and the data stream S0 in Example 6-8. Stream S0 has schema (c1 integer, c2 float). This example defines pattern C as C.c1 = 7. It defines firstc as first(C.c2). In other words, firstc will equal the value of c2 the first time c1 = 7. The query returns the relation in Example 6-9.

Example 6-7 FIRST Function Query

<query id="q9"><![CDATA[ 
    select 
        T.firstc,
        T.lastc,
        T.Ac1,
        T.Bc1,
        T.avgCc1, 
        T.Dc1 
    from 
        S0 
    MATCH_RECOGNIZE ( 
        MEASURES 
            first(C.c2) as firstc, 
            last(C.c2) as lastc, 
            avg(C.c1) as avgCc1, 
            A.c1 as Ac1, 
            B.c1 as Bc1, 
            D.c1 as Dc1 
        PATTERN(A B C* D) 
        DEFINE 
            A as A.c1 = 30, 
            B as B.c2 = 10.0, 
            C as C.c1 = 7, 
            D as D.c1 = 40
    ) as T
]]></query>

Example 6-8 FIRST Function Stream Input

Timestamp   Tuple
 1000       33,0.9
 3000       44,0.4
 4000       30,0.3
 5000       10,10.0
 6000        7,0.9
 7000        7,2.3
 9000        7,8.7
11000       40,6.6
15000       19,8.8
17000       30,5.5
20000        5,10.0
23000       40,6.6
25000        3,5.5
30000       30,2.2
35000        2,10.0
40000        7,5.5
44000       40,8.9

Example 6-9 FIRST Function Relation Output

Timestamp   Tuple Kind  Tuple
11000:      +           0.9,8.7,30,10,7.0,40
23000:      +           ,,30,5,,40
44000:      +           5.5,5.5,30,2,7.0,40

LAST

Syntax

Surrounding text describes last.gif.

Purpose

LAST returns the value of the specified stream element the last time the specified pattern is matched.

The type of the specified stream element may be any of:

The return type of this function depends on the type of the specified stream element.

This function takes a single argument made up of the following period-separated values:

See Also:

Examples

Consider the query q9 in Example 6-10 and the data stream S0 in Example 6-11. Stream S1 has schema (c1 integer, c2 float). This example defines pattern C as C.c1 = 7. It defines lastc as last(C.c2). In other words, lastc will equal the value of c2 the last time c1 = 7. The query returns the relation in Example 6-12.

Example 6-10 LAST Function Query

<query id="q9"><![CDATA[ 
    select 
        T.firstc,
        T.lastc,
        T.Ac1,
        T.Bc1,
        T.avgCc1, 
        T.Dc1 
    from 
        S0 
    MATCH_RECOGNIZE ( 
        MEASURES 
            first(C.c2) as firstc, 
            last(C.c2) as lastc, 
            avg(C.c1) as avgCc1, 
            A.c1 as Ac1, 
            B.c1 as Bc1, 
            D.c1 as Dc1 
        PATTERN(A B C* D) 
        DEFINE 
            A as A.c1 = 30, 
            B as B.c2 = 10.0, 
            C as C.c1 = 7, 
            D as D.c1 = 40
    ) as T
]]></query>

Example 6-11 LAST Function Stream Input

Timestamp   Tuple
 1000       33,0.9
 3000       44,0.4
 4000       30,0.3
 5000       10,10.0
 6000        7,0.9
 7000        7,2.3
 9000        7,8.7
11000       40,6.6
15000       19,8.8
17000       30,5.5
20000        5,10.0
23000       40,6.6
25000        3,5.5
30000       30,2.2
35000        2,10.0
40000        7,5.5
44000       40,8.9

Example 6-12 LAST Function Relation Output

Timestamp   Tuple Kind  Tuple
11000:      +           0.9,8.7,30,10,7.0,40
23000:      +           ,,30,5,,40
44000:      +           5.5,5.5,30,2,7.0,40

MAX

Syntax

Surrounding text describes max.gif.

Purpose

MAX returns maximum value of expr. Its datatype depends on the datatype of the argument.

Examples

Consider the query test_max_timestamp in Example 6-13 and the data stream S15 in Example 6-14. Stream S15 has schema (c1 int, c2 timestamp). The query returns the relation in Example 6-15.

Example 6-13 MAX Function Query

<query id="test_max_timestamp"><![CDATA[ 
    select max(c2) from S15[range 2]
]]></query>

Example 6-14 MAX Function Stream Input

Timestamp   Tuple
  10        1,"08/07/2004 11:13:48"
2000         ,"08/07/2005 11:13:48"
3400        3,"08/07/2006 11:13:48"
4700         ,"08/07/2007 11:13:48"
h 8000 
h 200000000

Example 6-15 MAX Function Relation Output

Timestamp   Tuple Kind  Tuple
   0:       + 
  10:       - 
  10:       +           08/07/2004 11:13:48
2000:       -           08/07/2004 11:13:48
2000:       +           08/07/2005 11:13:48
2010:       -           08/07/2005 11:13:48
2010:       +           08/07/2005 11:13:48
3400:       -           08/07/2005 11:13:48
3400:       +           08/07/2006 11:13:48
4000:       -           08/07/2006 11:13:48
4000:       +           08/07/2006 11:13:48
4700:       -           08/07/2006 11:13:48
4700:       +           08/07/2007 11:13:48
5400:       -           08/07/2007 11:13:48
5400:       +           08/07/2007 11:13:48
6700:       -           08/07/2007 11:13:48
6700:       +

MIN

Syntax

Surrounding text describes min.gif.

Purpose

MIN returns minimum value of expr. Its datatype depends on the datatype of its argument.

Examples

Consider the query test_min_timestamp in Example 6-16 and the data stream S15 in Example 6-17. Stream S15 has schema (c1 int, c2 timestamp). The query returns the relation in Example 6-18.

Example 6-16 MIN Function Query

<query id="test_min_timestamp"><![CDATA[ 
    select min(c2) from S15[range 2]
]]></query>

Example 6-17 MIN Function Stream Input

Timestamp   Tuple
  10        1,"08/07/2004 11:13:48"
2000         ,"08/07/2005 11:13:48"
3400        3,"08/07/2006 11:13:48"
4700         ,"08/07/2007 11:13:48"
h 8000 
h 200000000

Example 6-18 MIN Function Relation Output

Timestamp   Tuple Kind  Tuple
   0:       + 
  10:       - 
  10:       +           08/07/2004 11:13:48
2000:       -           08/07/2004 11:13:48
2000:       +           08/07/2004 11:13:48
2010:       -           08/07/2004 11:13:48
2010:       +           08/07/2005 11:13:48
3400:       -           08/07/2005 11:13:48
3400:       +           08/07/2005 11:13:48
4000:       -           08/07/2005 11:13:48
4000:       +           08/07/2006 11:13:48
4700:       -           08/07/2006 11:13:48
4700:       +           08/07/2006 11:13:48
5400:       -           08/07/2006 11:13:48
5400:       +           08/07/2007 11:13:48
6700:       -           08/07/2007 11:13:48
6700:       +

SUM

Syntax

Surrounding text describes sum.gif.

Purpose

SUM returns the sum of values of expr. This function takes as an argument any bigint, float, or integer expression. The function returns the same datatype as the numeric datatype of the argument.

Examples

Consider the query q3 in Example 6-19 and the data stream S1 in Example 6-20. Stream S1 has schema (c1 integer, c2 bigint). The query returns the relation in Example 6-21. For more information on range, see "Range-Based Stream-to-Relation Window Operators".

Example 6-19 SUM Query

<query id="q3"><![CDATA[ 
    select sum(c2) from S1[range 5]
]]></query>

Example 6-20 SUM Stream Input

Timestamp   Tuple
1000         5,
1000        10,5
2000          ,4
3000        30,6
5000        45,44
7000        55,3
h 200000000

Example 6-21 SUM Relation Output

Timestamp   Tuple Kind  Tuple
    0:      +
 1000:      -
 1000:      +           5
 2000:      -           5
 2000:      +           9
 3000:      -           9
 3000:      +           15
 5000:      -           15
 5000:      +           59
 6000:      -           59
 6000:      +           54
 7000:      -           54
 7000:      +           53
 8000:      -           53
 8000:      +           47
10000:      -           47
10000:      +           3
12000:      -           3
12000:      +

XMLAGG

Syntax

Surrounding text describes xmlagg.gif.

Purpose

XMLAGG returns a collection of XML fragments as an aggregated XML document. Arguments that return null are dropped from the result.

You can control the order of fragments using an order_by_clause. For more information, see Section 14.2.6, "Sorting Query Results".

Examples

This section describes the following XMLAGG examples:

XMLAGG and XMLELEMENT

Consider the query tkdata67_q1 in Example 6-22 and the input relation in Example 6-23. Stream tkdata67_S0 has schema (c1 integer, c2 float). This query uses XMLELEMENT to create XML fragments from stream elements and then uses XMLAGG to aggregate these XML fragments into an XML document. The query returns the relation in Example 6-24.

For more information about XMLELEMENT, see "xmlelement_expr".

Example 6-22 XMLAGG Query

<query id="tkdata67_q1"><![CDATA[ 
    select 
        c1, 
        xmlagg(xmlelement("c2",c2)) 
    from 
        tkdata67_S0[rows 10] 
    group by c1
]]></query>

Example 6-23 XMLAGG Relation Input

Timestamp           Tuple
 1000               15, 0.1
 1000               20, 0.14
 1000               15, 0.2
 4000               20, 0.3
10000               15, 0.04
h 12000

Example 6-24 XMLAGG Relation Output

Timestamp   Tuple Kind  Tuple
1000:       +           15,<c2>0.1</c2>
                           <c2>0.2</c2>
1000:       +           20,<c2>0.14</c2>
4000:       -           20,<c2>0.14</c2>
4000:       +           20,<c2>0.14</c2>
                           <c2>0.3</c2>
10000:      -           15,<c2>0.1</c2>
                           <c2>0.2</c2>
10000:      +           15,<c2>0.1</c2>
                           <c2>0.2</c2>
                           <c2>0.04</c2>

XMLAGG and ORDER BY

Consider the query tkxmlAgg_q5 in Example 6-25 and the input relation in Example 6-26. Stream tkxmlAgg_S1 has schema (c1 int, c2 xmltype). These query selects xmltype stream elements and uses XMLAGG to aggregate them into an XML document. This query uses an ORDER BY clause to order XML fragments. The query returns the relation in Example 6-27.

Example 6-25 XMLAGG and ORDER BY Query

<query id="tkxmlAgg_q5"><![CDATA[ 
    select 
        xmlagg(c2), 
        xmlagg(c2 order by c1) 
    from 
        tkxmlAgg_S1[range 2]
]]></query>

Example 6-26 XMLAGG and ORDER BY Relation Input

Timestamp           Tuple
1000                1, "<a>hello</a>"
2000               10, "<b>hello1</b>"
3000               15, "<PDRecord><PDName>hello</PDName></PDRecord>"
4000                5, "<PDRecord><PDName>hello</PDName><PDName>hello1</PDName></PDRecord>"
5000               51, "<PDRecord><PDId>6</PDId><PDName>hello1</PDName></PDRecord>"
6000               15, "<PDRecord><PDId>46</PDId><PDName>hello2</PDName></PDRecord>"
7000               55, "<PDRecord><PDId>6</PDId><PDName>hello2</PDName><PDName>hello3</PDName></PDRecord>"

Example 6-27 XMLAGG and ORDER BY Relation Output

Timestamp   Tuple Kind  Tuple
    0:      +
1000:       -
1000:       +           <a>hello</a>
                       ,<a>hello</a>
2000:       -           <a>hello</a>
                       ,<a>hello</a>
2000:       +           <a>hello</a>
                        <b>hello1</b>
                       ,<a>hello</a>
                        <b>hello1</b>
3000:       -           <a>hello</a>
                        <b>hello1</b>
                       ,<a>hello</a>
                        <b>hello1</b>
3000:       +           <b>hello1</b>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
                       ,<b>hello1</b>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
4000:       -           <b>hello1</b>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
                       ,<b>hello1</b>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
4000:       +           <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
5000:       -           <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDName>hello</PDName>
                        </PDRecord>
5000:       +           <PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
6000:       -           <PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDName>hello</PDName>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
6000:       +           <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>46</PDId>
                          <PDName>hello2</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDId>46</PDId>
                          <PDName>hello2</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
7000:       -           <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>46</PDId>
                          <PDName>hello2</PDName>
                        </PDRecord>
                       ,<PDRecord>
                          <PDId>46</PDId>
                          <PDName>hello2</PDName>
                        </PDRecord>
                        <PDRecord>
                          <PDId>6</PDId>
                          <PDName>hello1</PDName>
                        </PDRecord>
Go to previous page
Previous		 Go to next page
Next
Go to Documentation Home
Home		 Go to Book List
Book List		 Go to Table of Contents
Contents		 Go to Index
Index		 Go to Feedback page
Contact Us