CARTESIAN/CROSS JOIN

Hello
I use OBIEE version 11.1.1.6.0. I created made dummy table and column in the physical layer. and join all the dimension table with this fact. but I don't know how the business model to deal with it. can someone help me to create a cross join?
Thank you

Published by: 968086 on October 30, 2012 12:59 AM

Follow this link
http://www.rittmanmead.com/2009/08/Oracle-BI-EE-10-1-3-4-1-reporting-on-non-transactional-dimension-values-equivalence-of-outer-joins/

Score pls help if

Tags: Business Intelligence

Similar Questions

Difference between a CROSS JOIN and a Cartesian product of the noted comma?

Hello everyone,

Oracle version: Oracle Database 11 g Enterprise Edition Release 11.2.0.1.0 - 64 bit
OS: Linux Fedora Core 17 (x86_64)

I was practicing on recursive subquery factoring based on oracle examples available in the documentation
http://docs.Oracle.com/CD/E11882_01/server.112/e26088/statements_10002.htm#i2129904

I was working on an example that displays the hierarchy of each manager with related employees. Here's how

WITH tmptab(empId, mgrId, lvl) AS
(
    SELECT  employee_id, manager_id, 0 lvl
    FROM employees
    WHERE manager_id IS NULL
    UNION ALL
    SELECT  employee_id, manager_id, lvl+1
    FROM employees, tmptab
    WHERE (manager_id = empId)
)
SEARCH DEPTH FIRST BY mgrId SET order1
SELECT LPAD(' ', lvl * 3, ' ') || empId AS empId
FROM tmptab;

Which gives the desired result

EMPID
---------------------
100
   101
      108
      109
      110
      111
      112
      113
      200
      203
      204
      205
      206
   102
      103
      104
      105
      106
      107
   114
      115
      116
      117
      118
      119
   120
      125
      126
      127
      128
      180
      181
      182
      183
   121
      129
      130
      131
      132
      184
      185
      186
      187
   122
      133
      134
      135
      136
      188
      189
      190
      191
   123
      137
      138
      139
      140
      192
      193
      194
      195
   124
      141
      142
      143
      144
      196
      197
      198
      199
   145
      150
      151
      152
      153
      154
      155
   146
      156
      157
      158
      159
      160
      161
   147
      162
      163
      164
      165
      166
      167
   148
      168
      169
      170
      171
      172
      173
   149
      174
      175
      176
      177
      178
      179
   201
      202

107 rows selected.

SQL>

However, by chance, I noticed that if I put CROSS JOIN instead of put a comma between table names, the same query behaves differently.

In other words, if instead of writing

. . .
UNION ALL
    SELECT  employee_id, manager_id, lvl+1
    FROM employees, tmptab
    WHERE (manager_id = empId)

I am writing

. . .
UNION ALL
    SELECT  employee_id, manager_id, lvl+1
    FROM employees CROSS JOIN tmptab
    WHERE (manager_id = empId)

I get the following error message

ERROR at line 4:
ORA-32044: cycle detected while executing recursive WITH query

Any idea?
Correct me if I'm wrong, but I remember, oracle supports as many JOIN CROSSROADS notation for Cartesian product (vector product =). For example

SQL> WITH tmptab1 AS
  2  (
  3      SELECT 'a1' AS colval FROM DUAL UNION ALL
  4      SELECT 'a2' AS colval FROM DUAL UNION ALL
  5      SELECT 'a3' AS colval FROM DUAL
  6  ),
  7  tmptab2 AS
  8  (
  9      SELECT 'b1' AS colval FROM DUAL UNION ALL
 10      SELECT 'b2' AS colval FROM DUAL
 11  )
 12  SELECT t1.colval, t2.colval
 13  FROM tmptab1 t1 CROSS JOIN tmptab2 t2;

CO CO
-- --
a1 b1
a2 b1
a3 b1
a1 b2
a2 b2
a3 b2

6 rows selected.

SQL> LIST 13
 13* FROM tmptab1 t1 CROSS JOIN tmptab2 t2
SQL>
SQL>
SQL> CHANGE /CROSS JOIN/,
 13* FROM tmptab1 t1 , tmptab2 t2
SQL> 
SQL>
SQL> LIST
  1  WITH tmptab1 AS
  2  (
  3  SELECT 'a1' AS colval FROM DUAL UNION ALL
  4  SELECT 'a2' AS colval FROM DUAL UNION ALL
  5  SELECT 'a3' AS colval FROM DUAL
  6  ),
  7  tmptab2 AS
  8  (
  9  SELECT 'b1' AS colval FROM DUAL UNION ALL
 10  SELECT 'b2' AS colval FROM DUAL
 11  )
 12  SELECT t1.colval, t2.colval
 13* FROM tmptab1 t1 , tmptab2 t2
SQL> 
SQL> /

CO CO
-- --
a1 b1
a2 b1
a3 b1
a1 b2
a2 b2
a3 b2

6 rows selected.

SQL>

So if the two rated commas and CROSS JOIN have the same semantics, why do I get a cycle mentioned above cites recursive subquery factoring while the same query works pretty well with comma between table instead of CROSS JOIN names? Because if a cycle is detected (= current element ancestor) it means that the product with the CROSS JOIN notation produces duplicates which are absent in the result of the Cartesian product rated comma.

I would appreciate it if you could kindly shed some light.

Thanks in advance,

Kind regards
Dariyoosh

Hello

dariyoosh wrote:
... Oracle terminology could become really confusing. But once again, according to the online glossary, a Cartesian product is apparently regarded as a join
http://docs.Oracle.com/CD/E11882_01/server.112/e25789/glossary.htm?type=popup#CNCPT44493
>

There is no doubt that a Cartesian product (also called cross join) is a join. If loops in a WITH recursive clause are detected after completing the joins, but before other conditions apply, the relevant question here is "what are the requirements to join?
In the ANSI syntax, the distinction is always clear. Join conditions occur in the... Clause WE

SELECT  employee_id, manager_id, lvl + 1
FROM      employees
JOIN        tmptab          ON  (manager_id = empId)     -- Join condition
;

and other conditions occur in the WHERE (or HAVING or CONNECT BY) clause.

SELECT  employee_id, manager_id, lvl + 1
FROM            employees
CROSS JOIN         tmptab
WHERE  (manager_id = empId)     -- NOT a join condition
;

In the joins of the former, it seems to be the case that any condition involving 2 or more tables (or the + indicator of outer join) is a condtion of join:

SELECT  employee_id, manager_id, lvl + 1
FROM      employees
,         tmptab
WHERE  (manager_id = empId)     -- Join condition
;

Inline query vs cross join

Hi all
Two tables which have no relationship with each other. For example, an Employees table and a systemparameter with a startworktime table.
We need a query with the data in the two tables:
Get all employees and the startworktime (which is the same for everyone)
Which is cheaper: an inline query or a product Cartesian or crossjoin?
Inine:
Select name, function
(by selecting startworktime in systemparameter)
employees;
Cartesian product:
SELECT name, function, startwoime
rktfrom used
Cross join systemparameter;
Your opinion about this.

Both these do the same thing. I seriously doubt if we would have the benefits of performance on the other.

Kind regards

cross join?
I was asked if the veiw (select what is wrttien to make the view) has a cross join and follwing some ANSI compliance. So my question is... as table 3 below is used by alias tp and same table3 is used by alias tp2... commandeer that be considered as a cross join?

Sorry, I'm very new to SQL...
```
select col1, col2, col3 from table1 t, table2 t1, table3 tp
where
---
---
---
AND tp.phs_id =
                      (SELECT tp2.phs_id
                         FROM table3 tp2
                        WHERE tp2.id = tsp.id
                         AND ROWNUM < 2)
```
Published by: user11168115 on May 14, 2009 12:43
Hello (and welcome)
No, would not be a CROSS JOIN, since your alias tp2 JOINED tp (I guess you mean tp even if you have a TSP). A CROSS JOIN is where no JOIN condition at all is specified for a table (also known as the Cartesian product ), and if it's compatible ANSI SQL will actually be the words CROSS JOIN specified in the JOIN clause. For example:
```
SELECT *
   FROM table1
 CROSS JOIN table2;
```
Otherwise, non-compliant ANSI:
```
SELECT *
  FROM table1, table2;
```

Auto cross join

Hello.

Why Cross join on a table without specifying alias does not work properly ?

with
    t1 as (
      select 1 as q1, 2 as q2 from dual union 
      select 3, 4 from dual union 
      select 5, 6 from dual
  )
select *
  from t1
    cross join t1
order by 1
        Q1         Q2         Q1         Q2
---------- ---------- ---------- ----------
         1          2          1          2 
         1          2          1          2 
         1          2          1          2 
         3          4          3          4 
         3          4          3          4 
         3          4          3          4 
         5          6          5          6 
         5          6          5          6 
         5          6          5          6 

 9 rows selected

11.2.0.3.0 - 64 bit

2837285 wrote:

Ok.

Why 10 gr 2 we have error ORA-00918 (same self-join) and 11 g works, but the wrong result?

Why don't you ask Oracle via a support request, it's a bug.

You can't say he is working on 11g "but which gives bad result." Clearly if it is to give a wrong result then it doesn't work.

What happens on 11g is the fact that it runs without the exception of ambiguous column. There are a few known bugs in the way SQL ANSI has implemented compared to the regular SQL Oracle syntax. Oracle worked to correct.

I don't have 12 c to test, but it may have already been corrected in this version.

Extract multiple lines using the technique of the cross join

Hello.

Can someone suggest a method to return 3 lines of a query when otherwise would return only one line?

I am trying to reach the analog SQL logic to this topic-

SELECT x.foo, p.id, p.name FROM people p CROSS JOIN (SELECT 1 AS foo UNION ALL SELECT 2 UNION ALL SELECT 3) x;

Reason: I want a result three rows (n) force and use GROUP BY with case statements to create 3 levels summary.

Here is a simple SQL logical expression that works for my setup - OBI

SELECT
     "- Nx_CSDG0_Repair_Orders (Depot Repair Views)".Repair_Number saw_0,
     "- Nx_CSDG0_Repair_Orders (Depot Repair Views)".SR_Operating_Unit_Name saw_1
FROM
     "[Noetix-NoetixGlobalRepository] NoetixViews for Oracle Service"     
WHERE 
     ("- Nx_CSDG0_Repair_Orders (Depot Repair Views)".Repair_Number = '338246')

But I think that I can't do--

SELECT
     "- Nx_CSDG0_Repair_Orders (Depot Repair Views)".Repair_Number saw_0,
     "- Nx_CSDG0_Repair_Orders (Depot Repair Views)".SR_Operating_Unit_Name saw_1
FROM
     "[Noetix-NoetixGlobalRepository] NoetixViews for Oracle Service"
     CROSS JOIN (SELECT 1 AS foo UNION ALL SELECT 2 UNION ALL SELECT 3)
WHERE 
     ("- Nx_CSDG0_Repair_Orders (Depot Repair Views)".Repair_Number = '338246')

But what peut do?

Thank you.

-cs

Hi CSeelig,

The BI server uses the ANSI SQL standard. Then all the SQL that follows this specification in OBIEE will work.

I did an example with intensification:
http://gerardnico.com/wiki/dat/OBIEE/logical_sql/obiee_sql_densification

You will see a cross join to make a densification.

See you soon
Nico

Pivot + cross join

Hello

I wrote the following query, which works very well:

select * from
(
     select
          to_char(SCB_OPENTIME, 'YYYY-MM') as curr_date,
          SCB_TASK
     from EM_DATA_MSR_SC
)
pivot
(
     count(SCB_TASK) for curr_date in
     (
          '2011-01',
          '2011-02',
          '2011-03',
          '2011-04',
          '2011-05',
          '2011-06',
          '2011-07',
          '2011-08',
          '2011-09',
          '2011-10',
          '2011-11',
          '2011-12'
     )
);

Now, I need to apply a cross thereon join:

select * from
(
     select
          to_char(SCB_OPENTIME, 'YYYY-MM') as curr_date,
          SCB_TASK
     from EM_DATA_MSR_SC
     where
          EMSCG_STATUS = status.name
)
pivot
(
     count(SCB_TASK) for curr_date in
     (
          '2011-01',
          '2011-02',
          '2011-03',
          '2011-04',
          '2011-05',
          '2011-06',
          '2011-07',
          '2011-08',
          '2011-09',
          '2011-10',
          '2011-11',
          '2011-12'
     )
)
cross join
(
    select 0, 'Closed' as name from DUAL union
    select 1, 'Denied' from DUAL union
    select 2, 'Open' from DUAL
) status;

When I try to run this query, I get this error message:

"STATUS"."NAME": invalid identifier

However, SQL is supposed to have brought on a deep level table references.
I'm doing something wrong?

Hello

If this is what you want, then you don't want a cross join. The inner join in the view online with no name is the only reference to beaches you need.

with ranges as
(
    select 0, 0 as mini, 20 as maxi from DUAL union
    select 1, 21, 50 from DUAL union
    select 2, 51, 300 from DUAL
)
select * from
(
     select  r.mini,
          r.maxi,
          to_char (t.DAT, 'YYYY-MM')     as curr_date,
          t.NUM
     from      TEST     t
     join      RANGES     r     on      t.NUM >= r.mini
                    and      t.NUM <= r.maxi
)
pivot
(
     count(NUM) for curr_date in
     (
          '2011-01',
          '2011-02',
          '2011-03',
          '2011-04',
          '2011-05',
          '2011-06',
          '2011-07',
          '2011-08',
          '2011-09',
          '2011-10',
          '2011-11',
          '2011-12'
     )
)
ORDER BY  mini
;

Depending on your data and your needs, you might want an outer join, not an inner join, like this:

...     from          RANGES     r
     LEFT OUTER JOIN     TEST     t     on      t.NUM >= r.mini
                         and      t.NUM <= r.maxi

Published by: Frank Kulash, November 7, 2011 12:20
Alternative ADED of outer join

[8i] need help with full outer join combined with a cross join...

I can't understand how to combine a full outer join with a different type of join... is it possible?

Here are some create table and insert for examples of database:

CREATE TABLE     my_tab1
(     record_id     NUMBER     NOT NULL     
,     workstation     VARCHAR2(4)
,     my_value     NUMBER
     CONSTRAINT my_tab1_pk PRIMARY KEY (record_id)
);

INSERT INTO     my_tab1
VALUES(1,'ABCD',10);
INSERT INTO     my_tab1
VALUES(2,'ABCD',15);
INSERT INTO     my_tab1
VALUES(3,'ABCD',5);
INSERT INTO     my_tab1
VALUES(4,'A123',5);
INSERT INTO     my_tab1
VALUES(5,'A123',10);
INSERT INTO     my_tab1
VALUES(6,'A123',20);
INSERT INTO     my_tab1
VALUES(7,'????',5);


CREATE TABLE     my_tab2
(     workstation     VARCHAR2(4)
,     wkstn_name     VARCHAR2(20)
     CONSTRAINT my_tab2_pk PRIMARY KEY (workstation)
);

INSERT INTO     my_tab2
VALUES('ABCD','WKSTN 1');
INSERT INTO     my_tab2
VALUES('A123','WKSTN 2');
INSERT INTO     my_tab2
VALUES('B456','WKSTN 3');

CREATE TABLE     my_tab3
(     my_nbr1     NUMBER
,     my_nbr2     NUMBER
);

INSERT INTO     my_tab3
VALUES(1,2);
INSERT INTO     my_tab3
VALUES(2,3);
INSERT INTO     my_tab3
VALUES(3,4);

And, the results that I want to get:

workstation     sum(my_value)     wkstn_name     my_nbr1     my_nbr2
---------------------------------------------------------------
ABCD          30          WKSTN 1          1     2
ABCD          30          WKSTN 1          2     3
ABCD          30          WKSTN 1          3     4
A123          35          WKSTN 2          1     2
A123          35          WKSTN 2          2     3
A123          35          WKSTN 2          3     4
B456          0          WKSTN 3          1     2
B456          0          WKSTN 3          2     3
B456          0          WKSTN 3          3     4
????          5          NULL          1     2
????          5          NULL          2     3
????          5          NULL          3     4

I tried a number of different things, google my problem and no luck yet...

SELECT     t1.workstation
,     SUM(t1.my_value)
,     t2.wkstn_name
,     t3.my_nbr1
,     t3.my_nbr2
FROM     my_tab1 t1
,     my_tab2 t2
,     my_tab3 t3
...

So, what I want, it's a full outer join of t1 and t2 on workstation and a cross join of one with the t3. I wonder if I can't find examples of it online because it is not possible...

Note: I'm stuck dealing with Oracle 8i

Thank you!!

Hello

The query I posted yesterday is a little more complex that it should be.
My_tab2.workstation is unique, there is no reason to make a separate subquery as mt1. We can join my_tab1 to my_tab2 and get the SUM in a subquery.

SELECT       foj.workstation
,       foj.sum_my_value
,       foj.wkstn_name
,       mt3.my_nbr1
,       mt3.my_nbr2
FROM       (     -- Begin in-line view foj for full outer join
          SELECT        mt1.workstation
          ,        SUM (mt1.my_value)     AS sum_my_value
          ,        mt2.wkstn_name
          FROM        my_tab1   mt1
          ,        my_tab2   mt2
          WHERE        mt1.workstation     = mt2.workstation (+)
          GROUP BY   mt1.workstation
          ,        mt2.wkstn_name
                        --
                UNION ALL
                        --
          SELECT      workstation
          ,      0      AS sum_my_value
          ,      wkstn_name
          FROM      my_tab2
          WHERE      workstation     NOT IN (     -- Begin NOT IN sub-query
                                           SELECT      workstation
                                   FROM      my_tab1
                                   WHERE      workstation     IS NOT NULL
                                 )     -- End NOT IN sub-query
       ) foj     -- End in-line view foj for full outer join
,       my_tab3  mt3
ORDER BY  foj.wkstn_name
,       foj.workstation
,       mt3.my_nbr1
,       mt3.my_nbr2
;

Thanks for posting the CREATE TABLE and INSERT statements, and very clear expected results!

user11033437 wrote:
... So, what I want, it's a full outer join of t1 and t2 on workstation and a cross join of one with the t3.

She, exactly!
The trickiest part is when and how get SUM (my_value). You could address the question of exactly what my_tab3 must be attached to a cross that's exactly what should look like the result set of the full outer join between my_tab1 and my_tab2 to. To do this, take your desired results, remove columns that do not come from the outer join complete and delete duplicate rows. You will get:

workstation     sum(my_value)     wkstn_name
-----------     -------------   ----------
ABCD          30          WKSTN 1
A123          35          WKSTN 2
B456          0          WKSTN 3
????          5          NULL

So the heart of the problem is how to get these results of my_tab1 and my_tab2, which is done in the subquery FOJ above.

I tried to use auto-documenté in my code names. I hope you can understand.
I could spend hours explaining the different parts of this query more in detail, but I don't know that I would lose some of that time, explain things that you already understand. If you want an explanation of the specific element (s), let me know.

the group causing a Cartesian merge join?

Hi all

Database Version: Oracle Database 10g Enterprise Edition Release 10.2.0.5.0 - 64bi
Database Version: PL/SQL Release 10.2.0.5.0 - Production

I wrote a query that runs in less than a second and produced a few hundred lines (as you can see below). I wanted only to find the number of separate incidents ID returned so I put a group by at the end and it ran forever.

I realized a plan to explain (as shown below) to the query with the group inside and he tries to make a Cartesian join!

A WTF happening? I've never known a group by statement before.

basic structure of the tables is the following:

SQL> desc answer_master
 Name                                                                     Null?    Type
 ------------------------------------------------------------------------ -------- -----------------
 ANSWER_ID                                                                NOT NULL NUMBER(12)
 INCIDENT_ID                                                              NOT NULL NUMBER(12)
 PLAN_ID                                                                  NOT NULL NUMBER(12)
 QUESTION_ID                                                              NOT NULL NUMBER(12)

SQL> desc question_master
 Name                                                                     Null?    Type
 ------------------------------------------------------------------------ -------- -----------------
 QUESTION_PLAN_ID                                                         NOT NULL NUMBER(12)
 QUESTION_ID                                                              NOT NULL NUMBER(12)
 QUESTION_TYPE                                                            NOT NULL NUMBER(2)
 QUESTION_TEXT                                                            NOT NULL VARCHAR2(255)
 QUESTION_PARENT_ID                                                                NUMBER(12)

the SQL statement

SQL> explain plan for
  2  select incident_id
  3    from answer_master am
  4   where exists (select 'x'
  5                   from answer_master am1
  6                  where question_id in (select question_id
  7                               from question_master qms
  8                              where question_plan_id = 1477
  9                                and question_parent_id = 69067
 10                                and substr(question_text,-3) = 'PDF')
 11                    and am1.incident_id = am.incident_id)
 12    and exists (select 'x'
 13                   from answer_master am1
 14                  where question_id in (select question_id
 15                               from question_master qms
 16                              where question_plan_id = 1477
 17                                and question_parent_id = 69067
 18                                and substr(question_text,-3) != 'PDF')
 19                    and am1.incident_id = am.incident_id);

Explained.

Elapsed: 00:00:00.01
SQL> set linesize 132;
SQL> select plan_table_output from table(dbms_xplan.display('plan_table',null,'serial'));

PLAN_TABLE_OUTPUT                                                                                                                   
------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 1710257923                                                                                                         
                                                                                                                                    
----------------------------------------------------------------------------------------------------------                          
| Id  | Operation                        | Name                  | Rows  | Bytes | Cost (%CPU)| Time     |                          
----------------------------------------------------------------------------------------------------------                          
|   0 | SELECT STATEMENT                 |                       |     1 |    32 |   171   (6)| 00:00:01 |                          
|*  1 |  HASH JOIN RIGHT SEMI            |                       |     1 |    32 |   171   (6)| 00:00:01 |                          
|   2 |   VIEW                           | VW_SQ_2               |    71 |   923 |    70   (5)| 00:00:01 |                          
|   3 |    NESTED LOOPS                  |                       |    71 |  3266 |    70   (5)| 00:00:01 |                          
|*  4 |     TABLE ACCESS BY INDEX ROWID  | QUESTION_MASTER       |     1 |    35 |     2   (0)| 00:00:01 |                          
|*  5 |      INDEX RANGE SCAN            | QUESTION_MASTER_IX1   |     3 |       |     1   (0)| 00:00:01 |                          
|*  6 |     INDEX RANGE SCAN             | ANSWER_MASTER_QID_IX2 | 18731 |   201K|    68   (5)| 00:00:01 |                          
|   7 |   NESTED LOOPS                   |                       |   304 |  5776 |   100   (5)| 00:00:01 |                          
|   8 |    VIEW                          | VW_SQ_1               |    14 |   182 |    70   (5)| 00:00:01 |                          
|   9 |     HASH UNIQUE                  |                       |    14 |   644 |            |          |                          
|  10 |      NESTED LOOPS                |                       |    14 |   644 |    70   (5)| 00:00:01 |                          
|* 11 |       TABLE ACCESS BY INDEX ROWID| QUESTION_MASTER       |     1 |    35 |     2   (0)| 00:00:01 |                          
|* 12 |        INDEX RANGE SCAN          | QUESTION_MASTER_IX1   |     3 |       |     1   (0)| 00:00:01 |                          
|* 13 |       INDEX RANGE SCAN           | ANSWER_MASTER_QID_IX2 | 18731 |   201K|    68   (5)| 00:00:01 |                          
|* 14 |    INDEX RANGE SCAN              | ANSWER_MASTER_QID_IX3 |    22 |   132 |     2   (0)| 00:00:01 |                          
----------------------------------------------------------------------------------------------------------                          
                                                                                                                                    
Predicate Information (identified by operation id):                                                                                 
---------------------------------------------------                                                                                 
                                                                                                                                    
   1 - access("ITEM_2"="AM"."INCIDENT_ID")                                                                                          
   4 - filter("QUESTION_PLAN_ID"=1477 AND SUBSTR("QUESTION_TEXT",-3)!='PDF')                                                        
   5 - access("QUESTION_PARENT_ID"=69067)                                                                                           
   6 - access("QUESTION_ID"="QUESTION_ID")                                                                                          
  11 - filter("QUESTION_PLAN_ID"=1477 AND SUBSTR("QUESTION_TEXT",-3)='PDF')                                                         
  12 - access("QUESTION_PARENT_ID"=69067)                                                                                           
  13 - access("QUESTION_ID"="QUESTION_ID")                                                                                          
  14 - access("ITEM_1"="AM"."INCIDENT_ID")                                                                                          

33 rows selected.

Elapsed: 00:00:00.06
SQL> explain plan for
  2  select incident_id
  3    from answer_master am
  4   where exists (select 'x'
  5                   from answer_master am1
  6                  where question_id in (select question_id
  7                               from question_master qms
  8                              where question_plan_id = 1477
  9                                and question_parent_id = 69067
 10                                and substr(question_text,-3) = 'PDF')
 11                    and am1.incident_id = am.incident_id)
 12    and exists (select 'x'
 13                   from answer_master am1
 14                  where question_id in (select question_id
 15                               from question_master qms
 16                              where question_plan_id = 1477
 17                                and question_parent_id = 69067
 18                                and substr(question_text,-3) != 'PDF')
 19                    and am1.incident_id = am.incident_id)
 20  group by incident_id;

Explained.

Elapsed: 00:00:00.00
SQL> select plan_table_output from table(dbms_xplan.display('plan_table',null,'serial'));

PLAN_TABLE_OUTPUT                                                                                                                   
------------------------------------------------------------------------------------------------------------------------------------
Plan hash value: 1433543102                                                                                                         
                                                                                                                                    
-----------------------------------------------------------------------------------------------------------                         
| Id  | Operation                         | Name                  | Rows  | Bytes | Cost (%CPU)| Time     |                         
-----------------------------------------------------------------------------------------------------------                         
|   0 | SELECT STATEMENT                  |                       |     1 |    98 |    77   (6)| 00:00:01 |                         
|   1 |  HASH GROUP BY                    |                       |     1 |    98 |    77   (6)| 00:00:01 |                         
|   2 |   NESTED LOOPS                    |                       |     1 |    98 |    76   (4)| 00:00:01 |                         
|   3 |    NESTED LOOPS                   |                       |     1 |    87 |    74   (5)| 00:00:01 |                         
|   4 |     NESTED LOOPS                  |                       |     1 |    81 |    72   (5)| 00:00:01 |                         
|   5 |      MERGE JOIN CARTESIAN         |                       |     1 |    70 |     4   (0)| 00:00:01 |                         
|*  6 |       TABLE ACCESS BY INDEX ROWID | QUESTION_MASTER       |     1 |    35 |     2   (0)| 00:00:01 |                         
|*  7 |        INDEX RANGE SCAN           | QUESTION_MASTER_IX1   |     3 |       |     1   (0)| 00:00:01 |                         
|   8 |       BUFFER SORT                 |                       |     1 |    35 |     2   (0)| 00:00:01 |                         
|*  9 |        TABLE ACCESS BY INDEX ROWID| QUESTION_MASTER       |     1 |    35 |     2   (0)| 00:00:01 |                         
|* 10 |         INDEX RANGE SCAN          | QUESTION_MASTER_IX1   |     3 |       |     1   (0)| 00:00:01 |                         
|* 11 |      INDEX RANGE SCAN             | ANSWER_MASTER_QID_IX2 | 18731 |   201K|    68   (5)| 00:00:01 |                         
|* 12 |     INDEX RANGE SCAN              | ANSWER_MASTER_QID_IX3 |    22 |   132 |     2   (0)| 00:00:01 |                         
|* 13 |    INDEX RANGE SCAN               | ANSWER_MASTER_QID_IX2 |     1 |    11 |     2   (0)| 00:00:01 |                         
-----------------------------------------------------------------------------------------------------------                         
                                                                                                                                    
Predicate Information (identified by operation id):                                                                                 
---------------------------------------------------                                                                                 
                                                                                                                                    
   6 - filter("QUESTION_PLAN_ID"=1477 AND SUBSTR("QUESTION_TEXT",-3)='PDF')                                                         
   7 - access("QUESTION_PARENT_ID"=69067)                                                                                           
   9 - filter("QUESTION_PLAN_ID"=1477 AND SUBSTR("QUESTION_TEXT",-3)!='PDF')                                                        
  10 - access("QUESTION_PARENT_ID"=69067)                                                                                           
  11 - access("QUESTION_ID"="QUESTION_ID")                                                                                          
  12 - access("AM1"."INCIDENT_ID"="AM"."INCIDENT_ID")                                                                               
  13 - access("QUESTION_ID"="QUESTION_ID" AND "AM1"."INCIDENT_ID"="AM"."INCIDENT_ID")                                               

31 rows selected.

Elapsed: 00:00:00.00
SQL> spool off

Edit: change superiors--more/less to signs of! = to display the Forum

Published by: WhiteHat on February 11, 2011 09:37

In addition, if you are interested in rewriting so I think this might be an alternative (untested):

      select am1.incident_id
      from   answer_master am1
      ,      question_master qms
      where  am1.question_id        = qms.question_id
      and    qms.question_plan_id   = 1477
      and    qms.question_parent_id = 69067
      group by am1.incident_id
      having max(case when substr(qms.question_text,-3) = 'PDF' then 1 else 0 end) !=
             min(case when substr(qms.question_text,-3) = 'PDF' then 1 else 0 end);

Cross join to Hi in oracle

Hello

I have below the text values.

Platimum

Silver

VIP

Bronze

I want to get cross-from above like four

Platimum	Silver	VIP	Bronze
Platimum	VIP	Silver	Bronze
Platimum	Bronze	VIP	Silver
Platimum	Silver	VIP	Bronze

And so on. It should give me 16 combinations that I guess.

How can I achieve this?

Kind regards

Mahesh

Another way to generate permutations (if nocycle seems suspicious)

with

data in the form of

(select "Platimum" val of union double all the)

Select 'Cash' of all the double union

Select 'VIP' of all the double union

Select "Bronze" double

permutator (swap) as

(select val

from the data

Union of all the

Select swap. «, » || Val

data,.

permutator

where instr (permutation, val) = 0

)

Select the permutation

of the permutator

where regexp_count (permutation, ',') = 3

order by swapping

PERMUTATION
Bronze, Platimum, silver, VIP
Bronze, Platimum, VIP, Silver
Bronze, Silver, Platimum, VIP
Bronze, Silver, VIP, Platimum
Bronze, VIP, Platimum, Silver
Bronze, Silver, VIP, Platimum
Platimum, Bronze, silver, VIP
Platimum, Bronze, Silver, VIP
Platimum, silver, Bronze, VIP
Platimum, silver, VIP, Bronze
Platimum, VIP, Bronze, silver
Platimum, VIP, silver, Bronze
Silver, Bronze, Platimum, VIP
Silver, Bronze, VIP, Platimum
Money, Platimum, Bronze, VIP
Silver, Platimum, VIP, Bronze
Silver, VIP, Bronze, Platimum
Silver, VIP, Platimum, Bronze
VIP, Bronze, Platimum, silver
VIP Bronze, silver, Platimum
VIP, Platimum, Bronze, silver
VIP, Platimum, silver, Bronze
VIP, silver, Bronze, Platimum
VIP, money, Platimum, Bronze

MERGE JOIN Cartesian help
Hi all

I have a problem in writing the query below. The following query always shows a MERGE JOIN CARTESIAN in the plan to explain due to which the execution of the query takes a long time (almost 2 hours for 1,000 records).

I use "Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi. Please advice if I need to provide more information on that.
```
Select to_char(a.amnt_date,'DD-MON-YYYY') amnt_date, 
       SUM(a.AMNT_CON) amount
from amount a, 
     (SELECT area_no  
         FROM (SELECT * 
                  FROM area_usage 
                  WHERE area_hier_no = '201064') 
      CONNECT BY area_par_no = PRIOR area_no 
      START WITH area_no in (SELECT area_no 
                                FROM area 
                                WHERE area_code in ('EQ'))) b,
     (SELECT acc_no
         FROM (SELECT * 
                 FROM acc_usage 
                WHERE acc_hier_no = '1') 
      CONNECT BY acc_par_no = PRIOR acc_no 
      START WITH acc_no = 202917) c --  Account Hierarchy
where a.area_no = b.area_no 
   AND a.acc_no = c.acc_no
GROUP BY a.amnt_date
```
-----

PLAN_TABLE_OUTPUT

PLAN_TABLE_OUTPUT
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Hash value of plan: 2492729134

-----------------------------------------------------------------------------------------------------------------------------------
| ID | Operation | Name | Lines | Bytes | Cost (% CPU). Time | Pstart. Pstop |
-----------------------------------------------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | 692. 22144 | 3327 (32) | 00:00:40 |
| 1. HASH GROUP BY. 692. 22144 | 3327 (32) | 00:00:40 |
| 2. TABLE ACCESS BY LOCAL INDEX ROWID | AMOUNT | 138. 2898. 3326 (32) | 00:00:40 |
| 5: NESTED LOOPS | 692. 22144 | 3326 (32) | 00:00:40 |
| 4. THE CARTESIAN MERGE JOIN. 5. 55. 13 (0) | 00:00:01 |
| 5. VIEW | 1. 5. 5 (0) | 00:00:01 |
|* 6 | CONNECT BY WITH FILTERING |
| 7. TABLE ACCESS BY INDEX ROWID | ACC_USAGE | 1. 24. 2 (0) | 00:00:01 |
|* 8 | INDEX UNIQUE SCAN | ACU_PK | 1 | | 1 (0) | 00:00:01 |
| 9. NESTED LOOPS |
| 10. CONNECT PUMP |
| * 11 | TABLE ACCESS BY INDEX ROWID | ACC_USAGE | 1. 14. 5 (0) | 00:00:01 |
| * 12 | INDEX RANGE SCAN | ACU_AH_HAVE_AS_PARENT_FK_I | 4 | | 1 (0) | 00:00:01 |
| 13. KIND OF BUFFER. 5. 30. 13 (0) | 00:00:01 |
| 14. VIEW | 5. 30. 8 (0) | 00:00:01 |
| * 15 | CONNECT BY WITH FILTERING |
| * 16. TABLE ACCESS BY INDEX ROWID | AREA_USAGE |
| 17. NESTED LOOPS | 1. 32. 2 (0) | 00:00:01 |
| 18. TABLE ACCESS BY INDEX ROWID | AREA | 1. 21. 2 (0) | 00:00:01 |
| * 19. INDEX UNIQUE SCAN | AR_AR2_UK | 1 | | 1 (0) | 00:00:01 |
| * 20. INDEX UNIQUE SCAN | AU_PK | 8608. 94688 | 0 (0) | 00:00:01 |
| 21. NESTED LOOPS |
| 22. CONNECT PUMP |
| * 23. TABLE ACCESS BY INDEX ROWID | AREA_USAGE | 5. 85. 8 (0) | 00:00:01 |
| * 24. INDEX RANGE SCAN | ARU_APN_FKI | 8 | | 1 (0) | 00:00:01 |
| 25. RANGE OF PARTITION ALL THE | 1. 60.
| 26. CONVERSION OF BITMAP IN ROWID |
| 27. BITMAP AND |
| 28. CONVERSION OF BITMAP OF ROWID |
| * 29. INDEX RANGE SCAN | A_ARH_FK_I | 194K | 134 (7) | 00:00:02 | 1. 60.
| 30. CONVERSION OF BITMAP OF ROWID |
| * 31. INDEX RANGE SCAN | A_AH_FK_I | 194K | 488 (41) | 00:00:06 | 1. 60.
-----------------------------------------------------------------------------------------------------------------------------------

Information of predicates (identified by the operation identity card):
---------------------------------------------------

6 - access("ACC_USAGE".") ACC_PAR_NO "="ACC_USAGE"ADVANCE." ACC_NO")
8 - access("ACC_USAGE".") ACC_NO "= 202917 AND ACC_HIER_NO" = 1) «»
11 - filter ("ACC_HIER_NO" = 1)
12 - access("ACC_USAGE".") ACC_PAR_NO "="ACC_USAGE"ADVANCE." ACC_NO")
15 - access("AREA_USAGE".") AREA_PAR_NO "="AREA_USAGE"ADVANCE." AREA_NO')
16 - filter ("AREA_HIER_NO" = 201064)
19 - access ("AREA_CODE" = 'EQ')
20 - access ("AREA_HIER_NO"= "AREA_USAGE" AND 201064." AREA_NO "=" AREA_NO")
23 - filter ("AREA_HIER_NO" = 201064)
24 - access("AREA_USAGE".") AREA_PAR_NO "="AREA_USAGE"ADVANCE." AREA_NO')
29 - access("A".") AREA_NO '=' B '. ("' AREA_NO")
31 - access("A".") ACC_NO '=' C '. ("' ACC_NO")

54 selected lines.

Published by: Williams James Henderson on October 14, 2010 05:31
It is a very small difference, and I don't expect to cause huge changes in an execution plan. Either way to collect statistics, you would use the DBMS_STATS package.

Furthermore, your "formatted" execution plan is not exactly readable in your message.

Looking at your request a little more closely, I recommend you remove the SELECT * from subqueries. Something like that may be sufficient:
```
Select to_char(a.amnt_date,'DD-MON-YYYY') amnt_date,
       SUM(a.AMNT_CON) amount
from amount a,
     (SELECT area_no
         FROM (SELECT area_no
                    , area_par_no
                  FROM area_usage
                  WHERE area_hier_no = '201064')
      CONNECT BY area_par_no = PRIOR area_no
      START WITH area_no in (SELECT area_no
                                FROM area
                                WHERE area_code in ('EQ'))) b,
     (SELECT acc_no
         FROM (SELECT acc_no
                    , acc_par_no
                 FROM acc_usage
                WHERE acc_hier_no = '1')
      CONNECT BY acc_par_no = PRIOR acc_no
      START WITH acc_no = 202917) c --  Account Hierarchy
where a.area_no = b.area_no
   AND a.acc_no = c.acc_no
GROUP BY a.amnt_date
```
A SELECT statement using * and not at a later time using all the columns hides valuable information from Oracle. If you provide only the columns used Oracle may be able to find a better plan of execution.

Given that I don't know the structures of table if you can actually use all the columns. I still recommend the amendment because I think that it makes the code easier to understand.

is "join Cartesian" identical "merge join Cartesian"?
Hi guys,.

my plan to explain said the Cartesian merge join
[Oracle documentation | http://download-west.oracle.com/docs/cd/B10500_01/server.920/a96533/optimops.htm#45485] speaks of Cartesian join

Are these two the same?

Thank you
Hello

Both are the same but the NAP differs from 9i to 10g, since according to the algorithm basedon Version in! 0 g's
Kind of buffer.

Look at the example of wirh links.

http://www.juliandyke.com/optimisation/operations/MergeJoinCartesian.html
http://www.akadia.com/services/ora_interpreting_explain_plan.html

-Pavan Kumar N

Expected to a join behaviours?

Hello

formalities:

Select * from version of v$.

Oracle Database 11 g Enterprise Edition Release 11.2.0.3.0 - 64 bit Production

PL/SQL Release 11.2.0.3.0 - Production

CORE Production 11.2.0.3.0

AMT for 64-bit Windows: Version 11.2.0.3.0 - Production

NLSRTL Version 11.2.0.3.0 - Production

the question:

I was just wondering why this "join" between A and B do not work. A has no lines, B a 1 row

> my first answer... it should give 1 rank...

but alas... it doesn't!

the original request:

SELECT A.*,
  B.*,
CASE WHEN A.METH = 'PAL' THEN '1'
  ELSE A.CFGNR_OFRESTPAL 
END VWNR_OFRESTPAL
FROM    CDCROBOTVERDICHTLOCSTATVW A,
  CDCVERDICHTSETTINGSVW B

> No rows returned

so I checked the counts:

Select count (*) in the CDCROBOTVERDICHTLOCSTATVW A;

Select count (*) in the CDCVERDICHTSETTINGSVW B;

Hmmm... very strange

so I ran a few tests more:

SELECT  A.*, B.*, CASE WHEN A.METH = 'PAL' THEN '1' ELSE A.CFGNR_OFRESTPAL END VWNR_OFRESTPAL
FROM    CDCVERDICHTSETTINGSVW B , CDCROBOTVERDICHTLOCSTATVW A

> no line

SELECT A.*, B.*, CASE WHEN A.METH = 'PAL' THEN '1' ELSE A.CFGNR_OFRESTPAL END VWNR_OFRESTPAL
FROM    CDCVERDICHTSETTINGSVW B cross join CDCROBOTVERDICHTLOCSTATVW A

> no line

SELECT A.*, B.*, CASE WHEN A.METH = 'PAL' THEN '1' ELSE A.CFGNR_OFRESTPAL END VWNR_OFRESTPAL
FROM    CDCVERDICHTSETTINGSVW B full outer join CDCROBOTVERDICHTLOCSTATVW A on 1 =1

> This returns the line from B correctly

ARTNR	PROJECTNR	VARIANT	PROVIDER	QMAX	PRIO	METH	PALLETS	SOMQTY	FULLPALS	RESTQTY	RESTPALS	CFGNR_OFRESTPAL	FREELOCS	DUMMYRECIP	GEEN_VERDICHTING	MAXSRCPALS	MAXOPENPICKSETS	NR_OF_RESTPAL	ORDTYPE	PRIO_ARTNR	VERD_INRGRP	VWNR_OFRESTPAL
														VERDICHT		11	4	3	VERDICHT	QSE00918CLL80	CDCROB

SELECT A.*, B.*, CASE WHEN A.METH = 'PAL' THEN '1' ELSE A.CFGNR_OFRESTPAL END VWNR_OFRESTPAL
FROM  CDCVERDICHTSETTINGSVW B left join CDCROBOTVERDICHTLOCSTATVW A on 1 = 1

> it is also returns the same rank B

Does anyone have an idea on how to fix this... or where to look to fix this?

This is a cross join. It is a Cartesian product. If there are no rows in a table, then you get no rows in the result.

SQL> with a as (select 1 x from dual), b as (select 1 y from dual where 1=0)
  2  select a.*, b.*
  3  from a, b;                                                             

no rows selected

A left or right, or full outer join is apparently what you wanted.

Join the error in mapping!
Hello

I have 4 source tables and table a target. While the mapping of the source to the target table, I get following error:

Review: Default data set All stores data in a dataset must be attached. This data set has several sets of storage of data source disconnected. The sets are: [src1, src2, src3, src4]. To design a Cartesian product, create a join and select join cross in its properties.

Because of this error flow chart cannot be created and showing the error message.

Can someone tell me what exactly I have to do to make this work?

Thank you
Shri
If you need a Cartesian join between the tables, you must put a cross join. ODI is in need of a join (normal join or cross join) between all of the tables that you put in an interface.

Hint ordered and produced Cartesian
I have a question that needs to be tuned.

The initial request does not cross in plan product, when I change the order of the tables in the from clause (I placed the # fewer lines first and the more lines tables in the last) and I get product CARTESIAN MERGE JOIN where clause remaining the same, and the cost increases 7180 to 454 m.

This happens with ordered hint.
ranjus wrote:
I have a question that needs to be tuned.

The initial request does not cross in plan product, when I change the order of the tables in the from clause (I placed the # fewer lines first and the more lines tables in the last) and I get product CARTESIAN MERGE JOIN where clause remaining the same, and the cost increases 7180 to 454 m.

This happens with ordered hint.

Let's take a look at this with a simple test case. Suppose that:
* Table T1 contains 10,000 rows
* Table T2 contains 100 lines
* Table T3 contains 100,000 rows
* Table T4 contains 10 lines

You have the following query:
```
SELECT
  T1.C1,
  T2.C2,
  T3.C3,
  T4.C4
FROM
  T1,
  T2,
  T3,
  T4
WHERE
  T1.C1=T2.C2
  AND T2.C2=T3.C2
  AND T3.C3=T4.C3;
```
In the above, you can see that in the WHERE clause:
* Table T1 is attached to table T2
* The T2 table is attached to the tables T1 and T3
* Table T3 is attached to the T2 and T4 tables
* Table T4 joined table T3

It seems you are trying to do, it's something like this based on the number of rows in the tables:
```
SELECT /*+ ORDERED */
  T1.C1,
  T2.C2,
  T3.C3,
  T4.C4
FROM
  T4,
  T2,
  T1,
  T3
WHERE
  T1.C1=T2.C2
  AND T2.C2=T3.C2
  AND T3.C3=T4.C3;
```
In the light of the foregoing, we force optimizer Oracle to join the tables T4 and T2 - but there is no direct relationship between the two tables described in the WHERE clause. In one such case, Oracle optimizer must perform a Cartesian join between T4 and T2 tables. Then the T1 table has a relationship defined in the WHERE clause with the table T2, so that the join would not result in a Cartesian product. Finally, the T3 array has a relationship between tables T2 and T4, so that would not result in a Cartesian product.

Back to this example, suppose table T4 contained a single line. The execution plan still shows a Cartesian join, but this Cartesian join is no worse than a join of nested loops because there is only one line in one of the sources of the line.

In short, think what you try to force the optimizer to do. You cannot join always ascending rank cardinality estimates - it simply is not logic to try in some cases.

Charles Hooper
Co-author of "Expert Oracle practices: Oracle Database Administration of the Oak Table.
http://hoopercharles.WordPress.com/
IT Manager/Oracle DBA
K & M-making Machine, Inc.

CARTESIAN/CROSS JOIN

Similar Questions

Maybe you are looking for