Help!  -partition by clause

Similar Questions

• Request Cross-listing (Partition by Clause? Self-join?)

  Hello
  
  I need a query that cross-list will be a teacher teaches this semester of courses. Essentially, both fields must be the same (i.e.: CourseTitle & Section), while the third field is different (for example: subject).
  
  For example, Max Power is a Professor teaching the 3 course, one co-inscrit (123 ENG and JRL 123):
  

  LastName     FirstName     Subject     Section     CourseTitle
  Power          Max          ENG     123     English Composition
  Power          Max          ENG     452     Robert Frost Poetry     
  Power          Max          JRL     123     English Composition
  Power           Max          ENG      300     Faulkner & Twain

  The result of the desired query is:

  LastName     FirstName     Subject     Section     CourseTitle
  Power          Max          ENG     123     English Composition
  Power          Max          JRL     123     English Composition

  Basically, I don't need that the courses co-inscrit in the output. Is this an instance where I use a "Partition by Clause' or should I create a self-join?
  
  Well thanks for all help and comments.

  Hello

  user11137889 wrote:
  Unfortunately, I can't create tables. I don't have permission. I can't modify, add or remove all the data.

  It is no possible to work! You should have your own schema in a database of development, perhaps on your own machine, where you can create tables freely. You can develop or test software effectively without this ability.
  You must post CREATE TABLE and INSERT to your sample data, even if you can't test. Without this, all anyone can do is guess.

  If you want to only count each topic once, then replace COUNT (*) by COUNT (DISTINCT subject). Here, 'subject' can be almost any kind of expression, for example:

  count( DISTINCT substr ( ssbsect_ptrm_code
                          , 1
                   , 1
                   )
       ) over ( partition by  ssbsect_crse_numb
                    ,          scbcrse_title
           ) cnt
  

• How many fields we can add to the inside on the partition by clause?

  Hello world

  My version of DB is

  BANNER                                                         
  ----------------------------------------------------------------
  Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bi 
  PL/SQL Release 10.2.0.1.0 - Production                           
  CORE 10.2.0.1.0 Production                                         
  TNS for Linux: Version 10.2.0.1.0 - Production                   
  NLSRTL Version 10.2.0.1.0 - Production                           
  

  How many fields we can add to the inside on the partition by clause? I have added two and tried it works. So how can we add?

  Please, share your ideas.

  Kind regards

  BS2012.

  What is the purpose of this question? You can add as many as you have in the table.

  ----

  Ramin Hashimzade

• Have the help of the clause

  Oracle version 8i (we will migrate the db to 12 c in the next few months)

  This is defination and data table:

  create table sales

  (

  prod_id varchar2 (10),

  PROD_NAME varchar2 (30),

  salesman_id varchar2 (10),

  salesman_name varchar2 (30),

  commission_per varchar2 (10),

  commission_amt varchar2 (10)

  )

  /

  Insert in the value of sales ('prd1',' 1 ', ' 1', 'sm1' ", '10');

  Insert in the value of sales ('prd1',' 1 ', ' 2', 'sm2' ", '10');

  Insert in the value of sales ('prd1',' 1 ', ' 3', 'sm3' ", '10');

  Insert in the value of sales ('2 ', 'prd2', '1', 'sm1', '10.00' ");

  Insert in the value of sales ('2 ', 'prd2', '2', 'sm2', '10.00' ");

  Insert in the value of sales ('3 ', 'prd3', '2', 'sm2', '10.00' ");

  Insert in the value of sales ('3 ', 'prd3', '1', 'sm1', '15,00' ");

  Insert in the value of sales ('3 ', 'prd3', '3', 'sm3', '10.00' ");

  Insert in the value of sales ('4 ', 'prd4', '1', 'sm1', '10.00' ");

  Insert in the value of sales ('4', 'prd4', ' 2 ', 'sm2' ", '15');

  Insert in the value of sales ('5', 'prd5', ' 1 ', 'sm1' ", '15');

  Insert in the value of sales ('6', 'prd6', ' 1 ', 'sm1' ", '15');

  Insert in the value of sales ('6', 'prd6', ' 2 ', 'sm2' ", '15');

  Insert in the value of sales ('6', 'prd6', ' 3 ',' sm3 ', 15', ");

  Insert in the value of sales ('6', 'prd6', ' 4 ',' sm4 ', 15', ");

  Insert in the value of sales ('7', 'prd7', ' 1 ', 'sm1' ", '15');

  Insert in the value of sales ('7', 'prd7', ' 2 ', 'sm2' ", '15');

  Insert in the value of sales ('7', 'prd7', ' 3 ', 'sm3' ", '15');

  Insert in the value of sales ('7', 'prd7', ' 4 ', 'sm4' ", '15');

  Insert in the value of sales ('7', 'prd7', ' 7 ', 'sm7' ", '15');

  Insert in the value of sales ('7', 'prd7', ' 5 ', 'sm1' ", '10');

  Insert in the value of sales ('7', 'prd7', ' 6 ', 'sm6' ", '15');

  Dress uniform please:

  Unique Commision Amount/Commission Percentage:
  PROD_ID    PROD_NAME                      SALESMAN_I SALESMAN_NAME                  COMMISSION COMMISSION
  ---------- ------------------------------ ---------- ------------------------------ ---------- ----------
  1          prd1                           1          sm1                                       10
  2          prd2                           1          sm1                            10.00
  5          prd5                           1          sm1                                       15
  

  Means, we must identify the product ID for the business that gives the same percentage of commission sales amount or commission. Percentage of different Commission amount and the Commission:

  Different Commision Amount/Commission Percentage:
  PROD_ID    PROD_NAME                      SALESMAN_I SALESMAN_NAME                  COMMISSION COMMISSION
  ---------- ------------------------------ ---------- ------------------------------ ---------- ----------
  3          prd3                           2          sm2                            10.00
  3          prd3                           1          sm1                            15.00
  3          prd3                           3          sm3                            10.00
  4          prd4                           1          sm1                            10.00
  4          prd4                           2          sm2                                       15
  6          prd6                           1          sm1                                       15
  6          prd6                           2          sm2                                       15
  6          prd6                           3          sm3                            15
  6          prd6                           4          sm4                            15
  7          prd7                           1          sm1                                       15
  7          prd7                           2          sm2                                       15
  7          prd7                           3          sm3                                       15
  7          prd7                           4          sm4                                       15
  7          prd7                           7          sm7                                       15
  7          prd7                           5          sm1                                       10
  7          prd7                           6          sm6                                       15
  

  Means, for the highest of the product ID, assigned amount/percentage of different sales commission to sellers; that is for a product id, there is same amount/percentage of commission which would come at the first exit, if these lines of product codes should be in the second output. Kindly help me, since we use very old version, so I think that we cannot use analytics or new SQL functions.  So, please help me in a simple sql. Thank you.

  I hope that this works in version 8i.

  SQL > select prod_id
  2, prod_name
  3, salesman_id
  4, salesman_name
  5, commission_per
  6, commission_amt
  7 of)
  8. Select s.*
  9, mod (nvl (sum (commission_per) on (prod_id partition), 0), count (*) on (prod_id partition)) +.
  10 mod (nvl (sum (commission_amt) on (prod_id partition), 0), count (*) on (prod_id partition)) val
  11, row_number() over (partition by prod_id arrested by salesman_id) rno
  12 s sale
  13         )
  14 where val = 0
  15 and rno = 1;

  PROD_ID PROD_NAME SALESMAN_I SALESMAN_NAME COMMISSION COMMISSION
  ---------- ------------------------------ ---------- ------------------------------ ---------- ----------
  1          prd1                           1          sm1                                       10
  2          prd2                           1          sm1                            10.00
  5          prd5                           1          sm1                                       15

  SQL > select prod_id
  2, prod_name
  3, salesman_id
  4, salesman_name
  5, commission_per
  6, commission_amt
  7 of)
  8. Select s.*
  9, mod (nvl (sum (commission_per) on (prod_id partition), 0), count (*) on (prod_id partition)) +.
  10 mod (nvl (sum (commission_amt) on (prod_id partition), 0), count (*) on (prod_id partition)) val
  11, row_number() over (partition by prod_id arrested by salesman_id) rno
  12 s sale
  13         )
  14 where val! = 0 ;

  PROD_ID PROD_NAME SALESMAN_I SALESMAN_NAME COMMISSION COMMISSION
  ---------- ------------------------------ ---------- ------------------------------ ---------- ----------
  3          prd3                           2          sm2                            10.00
  3          prd3                           1          sm1                            15.00
  3          prd3                           3          sm3                            10.00
  4          prd4                           1          sm1                            10.00
  4          prd4                           2          sm2                                       15
  6          prd6                           2          sm2                                       15
  6          prd6                           1          sm1                                       15
  6          prd6                           4          sm4                            15
  6          prd6                           3          sm3                            15
  7          prd7                           7          sm7                                       15
  7          prd7                           5          sm1                                       10
  7          prd7                           4          sm4                                       15
  7          prd7                           3          sm3                                       15
  7          prd7                           6          sm6                                       15
  7          prd7                           1          sm1                                       15
  7          prd7                           2          sm2                                       15

  16 selected lines.

  SQL >

• help on the clause type

  I have a query as below
  
  Select empno, deptno, sum (sal) on tot_sal (partition deptno) from scott.emp;
  
  
  will give the result like this
  
  EMPNO DEPTNO TOT_SAL
  ---------- ---------- ----------
  7934 10 8750
  7782 10 8750
  7839 10 8750
  7902 20 10875
  7876 20 10875
  7566 20 10875
  7369 20 10875
  7788 20 10875
  
  8 selected lines.
  
  but as you can see the total sal and deptno repeats, what I want is wise dept total salary as shown below
  
  EMPNO DEPTNO TOT_SAL
  ---------- ---------- ----------
  7934
  7782
  7839
  Total for deptno 10 = 8750
  7902
  7876
  7566
  7369
  7788
  Total for deptno = 20 10875
  
  I want to use the type clause and get the above result

  970426 wrote:
  Karthik replied

  Published by: 970426 on January 24, 2013 04:13

  You are not yet able to determine good projection on your own?

  select
   nvl(to_char(empno),'Total for deptno '||deptno||': '||sal)
   empno
  from emp
  model
  partition by (deptno)
  dimension by (row_number() over(
                partition by deptno
                order by empno) rn)
  measures (sal, empno)
  rules(
  sal[0]=sum(sal) [any]
  )
  order by
  deptno, empno
  

  Katrick solution uses my idea of deptno [0]. It has a useless subquery, a useless measure, two unnecessary rules.
  Please don't give any of my points.
  Jiust learn a thing about it.
  Introduction to thinking on your own.

• Helps the model clause. Date field in the Dimension

  Hello, all.
  
  I'm having a problem with the clause type in my query. In my view, it may be a misunderstanding on my part, but I can't find any documentation that is opposed to the use of the date fields in the dimensions of my model.
  
  Given these data;
  

  LOGDATE               SHIFTNAME  BUILDING  UNIT  DOORNUMBER  INOUT    EVENTCOUNT
  2012-02-01 06:00:00      A        1800      R      Door 4     Out         14
  2012-02-01 06:00:00      A        1800      R      Door 4     In          15
  2012-02-01 18:00:00      D        1800      R      Door 4     Out         17
  2012-02-01 18:00:00      D        1800      R      Door 4     In          19
  2012-02-02 06:00:00      A        1800      R      Door 4     Out         14
  2012-02-02 06:00:00      A        1800      R      Door 4     In          14
  ...

  I write my SQL like this:
  

  SELECT logdate, shiftname, building, unit, doornumber, inout, eventcount
    FROM door_events
   WHERE building='1800' AND unit='R' AND doornumber=4
   MODEL DIMENSION BY (logdate, shiftname, building, unit, doornumber, inout)
         MEASURES (eventcount)
        RULES ()

  And I get the expected result. (The same table as above) But when I try to add rows using rules, I get no new line until the LOGDATE field is a date. For example, by using this code, I expect to get a new line, with a 100 EVENTCOUNT (and other fields as written) but I stiill get the same set of data (no line is added)
  

  SELECT logdate, shiftname, building, unit, doornumber, inout, eventcount
    FROM door_events
   WHERE building='1800' AND unit='R' AND doornumber=4
   MODEL DIMENSION BY (logdate, shiftname, building, unit, doornumber, inout)
         MEASURES (eventcount)
         RULES (eventcount[SYSDATE,'X','1800','R',4,'In'=100)

  However, when I turn the LOGDATE field in a CHAR data type (and make the SYSDATE a TANK in the same way) with the code below, all of a sudden I get the line I was expecting!
  

  SELECT TO_CHAR(logdate,'YYYY-MM-DD HH24:MI:SS') logdate, shiftname, building, unit, doornumber, inout, eventcount
    FROM door_events
   WHERE building='1800' AND unit='R' AND doornumber=4
   MODEL DIMENSION BY (logdate, shiftname, building, unit, doornumber, inout)
         MEASURES (eventcount)
         RULES (eventcount[TO_CHAR(SYSDATE,'YYYY-MM-DD HH24:MI:SS'),'X','1800','R',4,'In'=100)
  
  
  LOGDATE               SHIFTNAME  BUILDING  UNIT  DOORNUMBER  INOUT    EVENTCOUNT
  2012-02-01 06:00:00      A        1800      R      Door 4     Out         14
  2012-02-01 06:00:00      A        1800      R      Door 4     In          15
  2012-02-01 18:00:00      D        1800      R      Door 4     Out         17
  2012-02-01 18:00:00      D        1800      R      Door 4     In          19
  2012-02-02 06:00:00      A        1800      R      Door 4     Out         14
  2012-02-02 06:00:00      A        1800      R      Door 4     In          14
  ...
  2012-02-07 15:18:33      X        1800      R      Door 4     In          100

  Can someone explain this behavior? I intend to do some calculations on the LOGDATE dimension and I would like to keep it as a date. Thanks in advance for any help!

  I do not get the same result. When I use SYSDATE, I don't get the extra line. When I use TO_CHAR (SYDATE...) as you say, I still don't get the extra line.

  However, when I use a constant date, such as TO_DATE (' 2012/02/07 02:00 ',' YYYY-MM-DD HH24:MI:SS'), then I get the extra line.

  This behavior is described in http://docs.oracle.com/cd/E11882_01/server.112/e25554/sqlmodel.htm#BEIGGGFJ

  «Using UPSERT creates a new cell corresponding to that referenced on the left-hand side of the rule when the cell is absent, and the cell reference contains only positional references * qualified constants *.»

  Best regards, stew Ashton

  P.S. "Oracle Database 11g Enterprise Edition Release 11.2.0.2.0.

  Published by: stew Ashton on 7 February 2012 23:03

• Need help with where clause

  Hi all
  
  Could someone please help me on this one where clause.
  The structure of the table like this:
  
  < code >
  create table (test)
  primary key ID number,
  TYPE_ID number not null,
  name varchar2 (10));
  
  insert into test values (1, 10, 'name 1');
  insert into test values (2, 10, 'name 2');
  insert into test values (3, 20, 'name 3');
  insert into test values (4, 20, 'name 4');
  insert into test values (5, 20, 'name 5');
  insert into test values (6, 10, 'name 6');
  insert into test values (7, 10, 'name 7');
  commit;
  
  SQL * more > select * from test;
  
  TYPE_ID ID NAMES
  ---------- ---------- ----------
  1 10 name 1
  2 10 2 name
  3 20 name 3
  4 20 4 name
  5 20 5 name
  6 10 6 name
  7 10 7 name
  
  7 selected lines.
  < code >
  
  Here are the values of the drop-down list box which the user will be able to select
  0 all the project
  Type of project 10 10
  20 of the Type 20
  
  Based on the selection of the user, we need a SELECT to get all records or certain types of records.
  
  Thanks in advance!

  Hello

  If: selected_type_ID is the value in the drop-down list (0, 10, or 20), then:

  WHERE   :selected_type_id IN (test.type_id, 0)
  

  This assumes that test.type_id is never NULL.

• Help with the clause of "default".

  Hi all
  
  I have a scenario where there is a table with 75 columns. out of the 75 columns, 30 are varchar and 30 are digital. The remaining columns are other types of data when creating the table, that I forgot to mention the default values to varchar and numeric columns.
  
  Now, approximately 20,000 records are inserted into the table where some records contain NULL varchar and numeric. Now, I want to replace the null with varchar values and digital default values respectively.
  
  I can change the columns by adding the clause by default now and update the lines of NULL for the default values accordingly. But this process could be very heavy.
  
  Is there a better way to do this. Please help me.
  
  Thanks in advance,
  Rambeau

  I want to know that is it possible to set the default values for all columns of type varchar, both using a single SQL statement. Then you can update all the values NULL with default values.

  You are looking for something like this?

  SQL> create table t (empno default 1, ename default 'x') as select cast(null as integer) empno, cast(null as varchar2(10)) ename  from emp where rownum <= 4
  /
  Table created.
  
  SQL> select * from  t
  /
       EMPNO ENAME
  ---------- ----------
  
  4 rows selected.
  
  SQL> update t set empno = default, ename=default
  /
  4 rows updated.
  
  SQL> select * from  t
  /
       EMPNO ENAME
  ---------- ----------
           1 x
           1 x
           1 x
           1 x         
  
  4 rows selected.
  

• Help with Where clause

  Hello world

  I have an argument to function entry, lets call this file number. This argument can be 0 (zero) or any number like 9800. No negative number.

  My where clause, I would like to say, if the number is 0 (zero), then extract all the cases, otherwise just retrieve the lines/records that belong to the passed in file number.

  select * 
  from table1
  where case_num = (if case num is 0 then retrieve all cases, 
                  otherwise retrieve only the records that belong to the number passed in)
  

  Thanks in advance.

  Select *.

  FROM table1

  where case_num = num

  or num = 0;

  SY.

• need help on the clause type

  I have a temp table as below

  LOC X CNT

  20 20 C

  80 20 N

  60 30 C

  40 30 N

  100 40 N

  the rule is for each column of the LOC, there are 2 records of C, N and sum (NTC) must be = 100. If there is no value for C for a LOC (LOC = 40) then the value of the CNT corresponding to C should appear as null or 0 if the value of N is > 0.

  I need a model clause to do this, see below,

  LOC X CNT

  20 20 C

  80 20 N

  60 30 C

  40 30 N

  100 40 N

  0 TO 40 C

  
  

  
  

  Why why why? I am not able to understand the model clause. How did you learn model clause?

• With the help of the clause...

  Hello
  
  I try to use the With clause to preset some values in the table - as temporary structures. Currently, I need to create two of these tables called Param1 and Param2. I want this is to select than a Params array and make it have 2 fields. I get an error of parenthesis wrong however...
  
  
  Work with two "tables".
  -------------------
  with param1 as
  
  (select sum (stm.qty_rec) inco
  stock stm
  WHERE pnm_auto_key = 1
  GROUP BY pnm_auto_key
  ORDER BY pnm_auto_key),
  param2 as
  (select Select sum (saj.qty_adj)
  of stock_adjust saj
  join in-house stock stm on saj.stm_auto_key = stm.stm_auto_key
  parts_master MFN on stm.pnm_auto_key = pnm.pnm_auto_key inner join
  where pnm_auto_key = 1
  Group of pnm_auto_key)
  Select (INCO + SELECT) in param1, param2
  -------------------
  
  Does not not (with params), select (...)-> invalid table name
  ----------------
  with params as
  (select sum (stm.qty_rec) INCO
  stock stm
  WHERE pnm_auto_key = 1
  GROUP BY pnm_auto_key
  ORDER BY pnm_auto_key),
  (select sum (saj.qty_adj) SELECT
  of stock_adjust saj
  join in-house stock stm on saj.stm_auto_key = stm.stm_auto_key
  parts_master MFN on stm.pnm_auto_key = pnm.pnm_auto_key inner join
  where pnm_auto_key = 1
  Group of pnm_auto_key)
  Select params (INCO + SELECT)
  -----------------
  
  
  Does not (with params (select), (double) select...)-> lack of right parenthesis
  -------------------
  with params as
  (select (select sum (stm.qty_rec)
  stock stm
  WHERE pnm_auto_key = 1
  GROUP BY pnm_auto_key
  ORDER BY pnm_auto_key) INCO.
  (select sum (saj.qty_adj)
  of stock_adjust saj
  join in-house stock stm on saj.stm_auto_key = stm.stm_auto_key
  parts_master MFN on stm.pnm_auto_key = pnm.pnm_auto_key inner join
  where pnm_auto_key = 1
  (Pnm_auto_key group) SELECT DOUBLE)
  Select params (INCO + SELECT)
  -------------------

  for example

  with params as (select (select sum(stm.qty_rec)
                          from stock stm
                          WHERE pnm_auto_key = 1) INCO,
                         (select sum(saj.qty_adj)
                          from stock_adjust saj
                          inner join stock stm on saj.stm_auto_key = stm.stm_auto_key
                          inner join parts_master pnm on stm.pnm_auto_key = pnm.pnm_auto_key
                          where pnm_auto_key = 1) OUTG
                 FROM dual
                )
  select (INCO + OUTG) from params
  

• Help - &gt; Partition by report SQL

  Our users have a problem with the formula on the reports below. I don't know what that means. Our users have said that they need
  
  (Tons used (s) / spread miles) * 2000 lb/ton = thousand lbs/spread for their report.
  
  And I see below the formula used by our developer. Don't know how to make for good reading. Can you let me know the formula used in the code below could in fact represent the formula above. Users complained as below
  user-> column double and triple the amount, if there are multiple records for the same truck/driver/beat/with the same amount for used dry carpet and Miles to spread. If miles of spread or used dry carpet is different if calculated correctly.
  
  
  Published by: Lucy discover March 28, 2011 13:50

  Have you tried my suggestion above?

  (i.e. without using the SUM)

• [10g] need help with order by hierarchical query clause

  I have the following data samples:

  CREATE TABLE     bill_test1
  (     parent_part     CHAR(25)
  ,     child_part     CHAR(25)
  ,     line_nbr     NUMBER(5)
  ,     qty_per          NUMBER(9,5)
  );
  
  INSERT INTO bill_test1 VALUES ('ABC-1','ABC-10',100,1);
  INSERT INTO bill_test1 VALUES ('ABC-1','ABC-20',200,2);
  INSERT INTO bill_test1 VALUES ('ABC-1','ABC-30',300,3);
  INSERT INTO bill_test1 VALUES ('ABC-1','HARDWARE-1',401,10);
  INSERT INTO bill_test1 VALUES ('ABC-1','HARDWARE-2',402,5);
  INSERT INTO bill_test1 VALUES ('ABC-10','ABC-155',100,2);
  INSERT INTO bill_test1 VALUES ('ABC-10','HARDWARE-1',200,1);
  INSERT INTO bill_test1 VALUES ('ABC-155','RAW-2',100,4.8);
  INSERT INTO bill_test1 VALUES ('ABC-155','HARDWARE-3',200,3);
  INSERT INTO bill_test1 VALUES ('ABC-20','RAW-1',100,10.2);
  INSERT INTO bill_test1 VALUES ('ABC-30','RAW-3',100,3);

  And the following query gives me exactly what I want, in the order I want. However, I wonder if there is a way to get this order without creating the column SEQ, given that I don't need in my results

  SELECT     part_nbr
  ,     parent_part
  ,     child_part
  FROM     (
       SELECT     CONNECT_BY_ROOT b.parent_part                         AS part_nbr
       ,     b.parent_part
       ,     b.child_part
       ,     SYS_CONNECT_BY_PATH(b.line_nbr,' ')                    AS seq
       FROM     bill_test1 b
       ,     dual
       CONNECT BY     parent_part     = PRIOR child_part
       )
  WHERE          part_nbr     = 'ABC-1'
  ORDER BY     seq
  ;
  
  Results of above query, except with SEQ included in SELECT (just to show what I'm sorting off of):
  PART_NBR                     PARENT_PART                  CHILD_PART                   SEQ
  ---------------------------- ---------------------------- ---------------------------- -----------------
  ABC-1                        ABC-1                        ABC-10                        100
  ABC-1                        ABC-10                       ABC-155                       100 100
  ABC-1                        ABC-155                      RAW-2                         100 100 100
  ABC-1                        ABC-155                      HARDWARE-3                    100 100 200
  ABC-1                        ABC-10                       HARDWARE-1                    100 200
  ABC-1                        ABC-1                        ABC-20                        200
  ABC-1                        ABC-20                       RAW-1                         200 100
  ABC-1                        ABC-1                        ABC-30                        300
  ABC-1                        ABC-30                       RAW-3                         300 100
  ABC-1                        ABC-1                        HARDWARE-1                    401
  ABC-1                        ABC-1                        HARDWARE-2                    402

  Hello

  As long as there is that a single root, brothers and SŒURS of ORDER BY, you say, but you can not do in a subquery (well, you can, but usually there is no interest in a subquery). If the CONNECT BY in a subquery, there is no guarantee that the main request will preserve the hierarchical order which provides the subquery.

  The query you posted does not require a query of Tahina, so you can say:

  SELECT     CONNECT_BY_ROOT b.parent_part                         AS part_nbr
  ,     b.parent_part
  ,     b.child_part
  --,     SYS_CONNECT_BY_PATH(b.line_nbr,' ')                    AS seq
  FROM     bill_test1 b
  WHERE          CONNECT_BY_ROOT b.parent_part     = 'ABC-1'
  CONNECT BY     parent_part     = PRIOR child_part
  ORDER SIBLINGS BY     b.line_nbr
  ;
  

  I said the query you posted does not require a subquery. It also does not require double, so I guess that what you posted is a simplification of what you are really doing, and that could have a subquery. In particular, if you want to GROUP BY part_nbr, you need the subquery. We can use CONNECT_BY_ROOT expression in the WHERE clause (or, come to think of it, use a START WITH clause instead WHERE), but, for some reason, we cannot use CONNECT_BY_ROOT in a clause GROUP BY; We need to calculate CONNECT_BY_ROOT in a subquery, give it a name (like part_nbr) and Super GROUP OF this column in a query.

  This requires that there is that one node root. ORDER OF brothers and SŒURS means just that: children of a common parent will appear in the order, but the root nodes, which have no parents, may not be in order.

  Here's what I meant by using START WITH place WHERE:

  SELECT     CONNECT_BY_ROOT b.parent_part                         AS part_nbr
  ,     b.parent_part
  ,     b.child_part
  --,     SYS_CONNECT_BY_PATH(b.line_nbr,' ')                    AS seq
  FROM     bill_test1 b
  START WITH     b.parent_part     = 'ABC-1'
  CONNECT BY     parent_part     = PRIOR child_part
  ORDER SIBLINGS BY     b.line_nbr
  ;
  

  This should be much more effective, because it reduces the results before waste you your time by getting their descendants.
  With the help of a clause START WITH here is analogous to me sending you an email, saying "come to a meeting one my office at 03:00."
  By using a WHERE clause here is similar by sending me an e-mail to all members of society, saying: "come to a meeting one my office at 3:00" and then, as people get here, tell everyone except you could go back.

  Brothers and SŒURS ORDER BY was introduced in Oracle 9.

  Published by: Frank Kulash, December 9, 2010 14:39
  Added version with the START WITH clause

• How to use the clause type conditionally create new lines

  This question is just for learning - no object other than to start real world to create versions of some queries known to learn how to measure the performance of the MODEL clause clause of the MODEL versions.

  The question is: How can you write a STANDARD clause that pivots (updates) ONLY the lines that actually need to rotate and do NOT change the lines that have no value and cannot be rotated.

  I found a good site that has about two dozen example articles on different pieces. This first link is the beginning of the series.

  SQL features tutorials - Clause TYPE

  http://www.sqlsnippets.com/en/topic-11663.html

  This link is to a clause of FAQ of BluShadow MODEL version "how to convert rows to columns."

  Line - MODEL method string

  http://www.sqlsnippets.com/en/topic-11987.html

  The solution to this link uses this line of source:

  POSITION KEY VAL

  --- ---------- ----------

  R08 0 v1, v2, v3,.

  and this result set:

  POSITION KEY VAL

  --- ---------- ----------

  R08 1 v1

  R08 2 v2

  R08 3 v3

  The reason why it produces that result set is this clause of the statement of MODEL

  RETURN THE UPDATED LINES

  If you comment on this article, you will see that the original lines are all in position 0. This means that the solution reproduced EACH SOURCE LINE even if there is only one item in the list of values.

  Thus, even a line source with a single value (for example, the r01 'a') will be updated and updated this line is returned.

  For a large number of data sources with only a few lines that actually need to rotate it would be a great performance of infringement.

  This is the sample data source

  with t as (button 1, 'a' value 'abc' col2, col3 'def' of the double
  Select Union all 2, 'b', 'ghi', 'jkl' from dual
  Union all select 3, 'c, d, e, 'mno', 'pqr' from dual.
  Union all select 4, 'f', 'stu', 'vwx' from dual
  )
  Select the key, value, col2, col3
  t

  KEY, VALUE, COL2, COL3
  1, a, abc, def
  2, b, GHI, jkl
  3, "c, d, e, mno, pqr.
  4, f, stu, vwx

  Lines 1, 2 and 4 only have a value of (a, b, c respectively.

  Only line 3 must rotate. It contains "c, d, e" and which must become 3 lines in the result set

  Value of the key, col2 col3
  1 an abc def
  2B ghi jkl
  3 c mno, pqr
  3 d mno, pqr
  3 e mno, pqr
  4 f stu vwx

  The nut of the problem is that if the clause 'RETURN UPDATED ROWS' is used, then the lines 1, 2 and 4 must be updated in order to be returned in the result set. And if this clause is NOT used, then line 3 must be updated by 'c', d, e 'c' and two new product lines: one line for a ' and one for 'e '.

  How can you do this with the TYPE clause without the help of the clause "Return LINES UPDATE"? I don't have a solution to sample showing what I tried because I can't understand what it takes to even try.

  I have a solution that uses "RETURN UPDATED ROWS" but I want to compare this performance to the exercise, when this clause is NOT used.

  We ask BluShadow to add the solution of MODEL clause to this FAQ with other similar solutions.

  Like this?

  SQL > with t as
  () 2
  3 select the 1 key, 'a' val, 'abc' col2, col3 'def' of the double

  
  Select 4 Union all 2, 'b', 'ghi', 'jkl' from dual
  5 union all select 3, 'c, d, e, 'mno', 'pqr' from dual.
  Select 6 Union all the 4 'f', 'stu', 'vwx' from dual
  7)
  8. Select the key
  9, key_1
  10, val
  11, regexp_substr (val, ' [^,] +', 1, key_1) val_new
  12, col2
  13, col3
  14 t
  model 15
  16 partition by (key)
  dimension (1 key_1) 17
  18 measures (val, col2, col3, (length (val) - length (replace (val, ",")) + 1) as len)
  19 rules
  (20)
  21 val [for 1 to increment of len key_1 [1] 1] = val [1]
  22, col2 [for 1 to increment of len key_1 [1] 1] = col2 [1]
  23, col3 [for 1 to increment of len key_1 [1] 1] = col3 [1]
  24)
  25 order
  26 by key
  27, key_1;

  KEY KEY_1 VAL VAL_N NECK NECK
  ---------- ---------- ----- ----- --- ---
  1 1 has an abc def
  2 b 1 b ghi jkl
  3 1 c, d, e c mno pqr
  3 2 c, d, e d mno pqr
  3 3 c, d, e e mno pqr
  4 1 f f stu vwx

  6 selected lines.

• Help a resetting cumulative

  Hi, I need help with a query that has a running total which resets at certain points
  
  Version

  BANNER                                                                         
  --------------------------------------------------------------------------------
  Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production     
  PL/SQL Release 11.2.0.3.0 - Production                                           
  CORE  11.2.0.3.0  Production                                                         
  TNS for Linux: Version 11.2.0.3.0 - Production                                   
  NLSRTL Version 11.2.0.3.0 - Production 

  examples of data

  create table t ( stat varchar2(3)
                 , beg_dt date
                 , end_dt date
                 , dur    number);
  
  Insert into t  values ('ACT',to_date('01-JUL-2012','DD-MON-YYYY'),to_date('31-JUL-2012','DD-MON-YYYY'),31);
  Insert into t  values ('ACT',to_date('01-AUG-2012','DD-MON-YYYY'),to_date('31-AUG-2012','DD-MON-YYYY'),31);
  Insert into t  values ('ACT',to_date('01-SEP-2012','DD-MON-YYYY'),to_date('30-SEP-2012','DD-MON-YYYY'),30);
  Insert into t  values ('LOA',to_date('01-OCT-2012','DD-MON-YYYY'),to_date('25-OCT-2012','DD-MON-YYYY'),25);
  Insert into t  values ('LOA',to_date('26-OCT-2012','DD-MON-YYYY'),to_date('31-OCT-2012','DD-MON-YYYY'),6);
  Insert into t  values ('LOA',to_date('01-NOV-2012','DD-MON-YYYY'),to_date('04-NOV-2012','DD-MON-YYYY'),4);
  Insert into t  values ('ACT',to_date('05-NOV-2012','DD-MON-YYYY'),to_date('10-NOV-2012','DD-MON-YYYY'),6);
  Insert into t  values ('LOA',to_date('11-NOV-2012','DD-MON-YYYY'),to_date('30-NOV-2012','DD-MON-YYYY'),20);
  Insert into t  values ('LOA',to_date('01-DEC-2012','DD-MON-YYYY'),to_date('15-DEC-2012','DD-MON-YYYY'),15);
  Insert into t  values ('ACT',to_date('16-DEC-2012','DD-MON-YYYY'),to_date('31-DEC-2012','DD-MON-YYYY'),16);
  Insert into t  values ('LOA',to_date('01-JAN-2013','DD-MON-YYYY'),to_date('31-JAN-2013','DD-MON-YYYY'),31);
  Insert into t  values ('LOA',to_date('01-FEB-2013','DD-MON-YYYY'),to_date('05-FEB-2013','DD-MON-YYYY'),5);
  Insert into t  values ('LOA',to_date('06-FEB-2013','DD-MON-YYYY'),to_date('15-FEB-2013','DD-MON-YYYY'),10);
  Insert into t  values ('LOA',to_date('16-FEB-2013','DD-MON-YYYY'),to_date('25-FEB-2013','DD-MON-YYYY'),10);
  Insert into t  values ('LOA',to_date('26-FEB-2013','DD-MON-YYYY'),to_date('28-FEB-2013','DD-MON-YYYY'),3);
  Insert into t  values ('LOA',to_date('01-MAR-2013','DD-MON-YYYY'),to_date('10-MAR-2013','DD-MON-YYYY'),10);
  Insert into t  values ('ACT',to_date('11-MAR-2013','DD-MON-YYYY'),to_date('31-MAR-2013','DD-MON-YYYY'),21);
  Insert into t  values ('ACT',to_date('01-APR-2013','DD-MON-YYYY'),to_date('30-APR-2013','DD-MON-YYYY'),30);
  Insert into t  values ('LOA',to_date('01-MAY-2013','DD-MON-YYYY'),to_date('25-MAY-2013','DD-MON-YYYY'),25);
  Insert into t  values ('LOA',to_date('26-MAY-2013','DD-MON-YYYY'),to_date('31-MAY-2013','DD-MON-YYYY'),6);
  Insert into t  values ('LOA',to_date('01-JUN-2013','DD-MON-YYYY'),to_date('04-JUN-2013','DD-MON-YYYY'),4);
  Insert into t  values ('ACT',to_date('05-JUN-2013','DD-MON-YYYY'),to_date('30-JUN-2013','DD-MON-YYYY'),26);

  It's close.
  

  select stat
      , beg_dt
      , end_dt                             
      , dur
      , sum( case stat when 'LOA' then dur else 0 end ) over ( partition by stat order by beg_dt) tot1
   from t
  order by beg_dt
  
  STAT BEG_DT      END_DT      DUR TOT1
  ---- ----------- ----------- --- ----
  ACT  01-JUL-2012 31-JUL-2012  31    0 
  ACT  01-AUG-2012 31-AUG-2012  31    0 
  ACT  01-SEP-2012 30-SEP-2012  30    0 
  LOA  01-OCT-2012 25-OCT-2012  25   25 
  LOA  26-OCT-2012 31-OCT-2012   6   31 
  LOA  01-NOV-2012 04-NOV-2012   4   35 
  ACT  05-NOV-2012 10-NOV-2012   6    0 
  LOA  11-NOV-2012 30-NOV-2012  20   55 
  LOA  01-DEC-2012 15-DEC-2012  15   70 
  ACT  16-DEC-2012 31-DEC-2012  16    0 
  LOA  01-JAN-2013 31-JAN-2013  31  101 
  LOA  01-FEB-2013 05-FEB-2013   5  106 
  LOA  06-FEB-2013 15-FEB-2013  10  116 
  LOA  16-FEB-2013 25-FEB-2013  10  126 
  LOA  26-FEB-2013 28-FEB-2013   3  129 
  LOA  01-MAR-2013 10-MAR-2013  10  139 
  ACT  11-MAR-2013 31-MAR-2013  21    0 
  ACT  01-APR-2013 30-APR-2013  30    0 
  LOA  01-MAY-2013 25-MAY-2013  25  164 
  LOA  26-MAY-2013 31-MAY-2013   6  170 
  LOA  01-JUN-2013 04-JUN-2013   4  174 
  ACT  05-JUN-2013 30-JUN-2013  26    0 

  I need a total of stat LOA who is recovering with each Act break to get me:

  STAT BEG_DT      END_DT      DUR TOT1
  ---- ----------- ----------- --- ----
  ACT  01-JUL-2012 31-JUL-2012  31    0 
  ACT  01-AUG-2012 31-AUG-2012  31    0 
  ACT  01-SEP-2012 30-SEP-2012  30    0 
  LOA  01-OCT-2012 25-OCT-2012  25   25 
  LOA  26-OCT-2012 31-OCT-2012   6   31 
  LOA  01-NOV-2012 04-NOV-2012   4   35 
  ACT  05-NOV-2012 10-NOV-2012   6    0 
  LOA  11-NOV-2012 30-NOV-2012  20   20 
  LOA  01-DEC-2012 15-DEC-2012  15   35 
  ACT  16-DEC-2012 31-DEC-2012  16    0 
  LOA  01-JAN-2013 31-JAN-2013  31   31 
  LOA  01-FEB-2013 05-FEB-2013   5   36 
  LOA  06-FEB-2013 15-FEB-2013  10   46 
  LOA  16-FEB-2013 25-FEB-2013  10   56 
  LOA  26-FEB-2013 28-FEB-2013   3   59 
  LOA  01-MAR-2013 10-MAR-2013  10   69 
  ACT  11-MAR-2013 31-MAR-2013  21    0 
  ACT  01-APR-2013 30-APR-2013  30    0 
  LOA  01-MAY-2013 25-MAY-2013  25   25 
  LOA  26-MAY-2013 31-MAY-2013   6   31 
  LOA  01-JUN-2013 04-JUN-2013   4   35 
  ACT  05-JUN-2013 30-JUN-2013  26    0 

  Any help would be appreciated, thanks.

  Hello

  So you want a total separate whenever there is an "Act". It is what PARTITION BY. You must add the number of acts already encountered (which I called act_cnt) to the PARTITION BY clause that gets you the SUM.
  We can get act_cnt using the analytical functions such as COUNT or SUM, but, given an analytic function cannot be dependent on another calculated the same query, we need to use a subquery to calculate act_cnt.

  Here's one way:

  WITH     got_act_cnt     AS
  (
       SELECT     stat, beg_dt, end_dt, dur
       ,     COUNT (CASE WHEN stat = 'ACT' THEN 1 END)
                OVER (ORDER BY  beg_dt)   AS act_cnt
       FROM    t
  )
  SELECT       stat, beg_dt, end_dt, dur
  ,       SUM (CASE stat WHEN 'LOA' THEN dur ELSE 0 END)
           OVER ( PARTITION BY  stat
                   ,                act_cnt     -- ***** NEW *****
               ORDER BY      beg_dt
             )       AS totl
  FROM      got_act_cnt
  ORDER BY  beg_dt
  ;
  

  Are 'Act' and 'AOL' the only possible values for the stat? If so, the query above can probably be simplified.