SQL & PL/SQL :: Windowing Clause In Analytical Function?
Oct 5, 2010
I need to calculate the sum of values over a period of exactly one month (including the current row). Now if I use a windowing clause of "range between interval '1' month preceding and current row", the total period length is 1 month plus one day (being the day in the current record).
Basically, I want to sum over a period starting at "add_months(startdate, -1) + 1" up until startdate of each row.
drop table window_tst;
create table window_tst
( id number primary key
[Code]....
So instead of having 01-feb going back to 01-jan, it should only include 02-jan till 01-feb
I could of course recalculate the period length back to a number of days for each row, but that is not really what I would prefer, as it would make the code rather unreadable.
I want to use Analytical function instead of group by clause for below query..
select CASE WHEN ADMT.SOURCESYSTEM ='CLU' THEN COUNT(ADMT.TOTAL_COUNT)*5 ELSE COUNT(ADMT.TOTAL_COUNT) END TOTAL_COUNT from ESMARTABC.ABC_DRVR_MFAILS_TMP ADMT group by ADMT.SOURCESYSTEM
I'm posting below test case in which I'm not able to understand output for LAST_VALUE function. I'm expecting maximum value for the salary in a department. Because I'm partitioning by department and ordering a partition as assending so being last value it should give me maximum value within a partition i.e. department in this case.
CREATE TABLE TEST1 (AGG_DATE DATE, COL1 NUMBER(9), COL2 NUMBER(9), COL3 NUMBER(9)); Here is the test-data population script: insert into TEST1 (AGG_DATE, COL1, COL2, COL3) values (to_date('01-01-2012', 'dd-mm-yyyy'), 1, 1, 1); [code]....
The problem is when I wrote an analytical query, it is giving the BEGIN_DATE and END_DATE by taking all the partition values together and so instead of the values above, it is creating an answer as follows:
how to delete duplicated records from a table without using row_id. I found the duplicated rows from a table using Analytical Function. But i could not use the Analytical function in the where condition.
I have a function that returns the total sum of an account. From reports I call the function passing the account code. The function sums the values for that specific account code and returns the value. In my function I have the following code :
where account_code = P_CODE.
Eg. The value of :P_CODE is 'CS'.
I now want to pass multiple account codes ('CS','TV',LJ') to the function. How do I change the IN clause in the function to accommodate multiple values.
I have tried using the instr function, but it does not work. eg. AND instr(o.ACCOUNT_CODES,','||P_CODE||',') > 0
DECLARE cnt number(10); BEGIN SELECT COUNT(*) INTO CNT FROM TBL_ADDRESS WHERE ADDRESS_ZIP IN (SELECT * FROM TABLE(MY_PACK.STR2TBL('46227'))); DBMS_OUTPUT.PUT_LINE (cnt); END;
MY_PACK.STR2TBL() is a function which takes '|' delimited string, extracts values and returns a table of zipcodes. The function works fine and returns 46227 but the count returned is 0 instead of 280(count returned by replacing inner select with '46227').
can i create the user-defined functions and use them in the FILTER clause in the sem_match function? there are some built-in functions for the FILTER clasue. however, only one function (DATATYPE(literal)) support for date/time in the built-in functions. i want to implement some user-defined funcitons in the FILTER clause which can check time intervals in ontology. there are some functions about valid time in the WorkSpace Manager such as WM_OVERLAPS, WM_CONTAINS,WM_MEETS, etc. so, can i write some functions using the these valid time functions in WM and use them in the FILTER clause?
I was reading a tutorial for analytical function and i found something like this
sum(princial) keep(dense_rank first order by d_date) over partition by (userid, alias, sec_id, flow, p_date)
How to translate this into simple queries / subquery? i am aware that analytical function are faster but i would like to know how this can translate to using query without analytical function.
Is there any way to apply the restriction on analytical functions, just like WHERE and HAVING .AS we know that we can apply the restriction on table by using WHERE and grouping functions by using HAVING clause .
For Ex: Departments wise count including all employees record :
SQL> select count(*) over(partition by deptno) dept_Count, ce.* 2 from scott.emp ce 3 order by deptno, job;
DEPT_COUNT EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO ---------- ----- ---------- --------- ----- ----------- --------- --------- ------ 3 7934 MILLER CLERK 7782 1/23/1982 1300.00 10 3 7782 CLARK MANAGER 7839 6/9/1981 2450.00 10 3 7839 KING PRESIDENT 11/17/1981 5000.00 10 5 7788 SCOTT ANALYST 7566 4/19/1987 3000.00 20 [code]....
Now i am trying to get the max(Value_2011) keep (dense_rank Last order by Month_ID) but i get a NULL. I can understand its because the Month_ID accomodates all years but i only need it to look at Month_ID for 2011 and return me the last dense_rank value, how can i achieve this?
I tried a couple of different methods like Last_Value() but i have group by in my original statement and i think analytical functions dont like GROUP by if they are not part of it. How can i achieve this?
I need to calcaulate the salary avarage for three days prior, leaving the current row. That should happen to every row moving back words.I have given all the details.
create table Employee( ID VARCHAR2(4 BYTE) NOT NULL, name varchar(20), Start_Date DATE, Salary Number(8,2), mv_avg number(8,2) [code]....
My table has two date columns EFF_DT which is the start date and TERM_DT is the end date. The EFF_DT of the next record should be the next date of the TERM_DT record.
In the fourth record, the effective date should be 1-Oct-13 which is the next date to the last TERM_DT 30-Sep-13.As the is the break in the date, the output should show 15-Oct-13 sa the second start date.
Note: Refer to the PI_ID columns, there is a break in the date for the sale PI_ID 'ABC'.
Here I am trying to generate a pseudo column, so that the table with the pseudo column looks like as shown below. and I can use first_value and LAST_value by partitioning on the pseudo column to get the desired output.
1) CNT_VAL is the pseudo column: ----------------------------- CK_IDPI_IDEFF_DT TERM_DT CNT_VAL Mem1ABC1-Jan-1331-Mar-131 Mem1ABC1-Apr-1331-May-131 Mem1ABC1-Jun-1330-Sep-131
[code].....
My Query : ----------
I not getting the desired output here as the value in pseudo column is 3.
select CK_ID, PI_ID,EFF_DT,TERM_DT, (case when case_CONT - LAG(case_CONT,1) over (ORDER BY EFF_DT) = 0 then to_char(case_CONT) when case_CONT - LAG(case_CONT,1) over (ORDER BY EFF_DT) <> 0 then to_char(LAG(case_CONT,1) over (ORDER BY EFF_DT) + 1) else to_char(nvl(case_CONT,0))
We've got a query which returns one row, but uses an IN statement. The IN statement links to more than one row in the subquery. When we use a combination of DISTINCT and an ANALYTICAL sum, the sum total is multiplied by the number of rows in the sub query. Remove the DISTINCT and we get a single value.
A simplified example of the problem is below.
I can't see how a query which returns a single row then returns multiple values with the addition of a DISTINCT. Removing the analytical sum also provides a single row, but we need this in the actual query we're running. So it seems to be some combination of DISTINCT, ANALYTICAL SUM and IN query is causing multiple values to be returned.
CREATE TABLE go_test_distinct1 (gtd_value NUMBER); -- Three identical values -- To replicate the three identical values returned by
I am having a table with 5 lakhs transactions. I want to fetch the last balance for a particular date. So i have have returned a query like below.
SELECT curr_balance FROM transaction_details WHERE acct_num = '10'
[Code]...
This has to be executed for incrementing of 12 months to find the last balance for each particular month. But this query is having more cpu cost, 12 times it is taking huge time. how to remodify athe above query to get the results in faster way using analytical query. Whether this can be broken into two part in PL/SQL to achive the performance. ?
I have a exp .dmp file of a database that has over 1000 tables. I would like to import all tables that don't begin with 'C'. Is this possible with IMP (without listing each table I want)?
