Need help with analytical function (LAG)
The requirement is as I have a table with described columscol1 County flag Flag2
ABC 1 Y Y
XYZ 1 Y Y
XYZ 1 O NULL
xyz *2* N N
XYZ 2 Y NULL
DEF 1 Y Y
DEF 1 N NULL
To get the columns Flag2
1 assign falg2 as indicator for rownum = 1
2 check the colm1, count of current line with colm1, Earl of the previous line. The colm1 and the NTC are identical, should assign null...
Here's the query I used to get the values of Flag2
SELECT colm1, count, flag
BOX WHEN
LAG(Count, 1,null) OVER (PARTITION BY colm1 ORDER BY colm1 DESC NULLS LAST) IS NULL
and LAG(flag, 1, NULL) PLUS (SCORE FROM colm1 ORDER BY colm1, cycle DESC NULLS LAST) IS NULL
THEN the flag
END AS Flag2
FROM table1
but the query above returns the o/p below which is false
col1_ County flag Flag2
ABC 1 Y Y
XYZ 1 Y Y
XYZ 1 O NULL
xyz *2* N NULL
XYZ 2 Y NULL
DEF 1 Y Y
DEF 1 N NULL
Thank you
Published by: user9370033 on April 8, 2010 23:25
Well, you have not enough explained your full requirement in this
1 assign falg2 as indicator for rownum = 1
2 check the colm1, count of current line with colm1, Earl of the previous line. The colm1 and the NTC are identical, should assign null...
as you say not what Flag2 must be set on if com1 and cnt are not the same as the previous row.
But how about this as my first guess what you mean...
SQL> with t as (select 'abc' as col1, 1 as cnt, 'Y' as flag from dual union all
2 select 'xyz', 1, 'Y' from dual union all
3 select 'xyz', 1, 'Y' from dual union all
4 select 'xyz', 2, 'N' from dual union all
5 select 'xyz', 2, 'Y' from dual union all
6 select 'def', 1, 'Y' from dual union all
7 select 'def', 1, 'N' from dual)
8 -- END OF TEST DATA
9 select col1, cnt, flag
10 ,case when lag(col1) over (order by col1, cnt) is null then flag
11 when lag(col1) over (order by col1, cnt) = col1 and
12 lag(cnt) over (order by col1, cnt) = cnt then null
13 else flag
14 end as flag2
15 from t
16 /
COL CNT F F
--- ---------- - -
abc 1 Y Y
def 1 Y Y
def 1 N
xyz 1 Y Y
xyz 1 Y
xyz 2 Y Y
xyz 2 N
7 rows selected.
SQL>
Tags: Database
Similar Questions
-
Need help with the function or metric derivative to calculate percentages of threshold for a measure
Hi, first post to the community that I am a n00b Foglight needing help.
A thing (in fact the only thing) I like Microsoft SCOM is how this graph of the availability of a metric, and I want to do the same thing in Foglight. I understand that this could be a derived measure or a function, I need, but am a bit lost right now.
Let's say I have a metric and created thresholds as follows: normal included 0, 50 inclusive, warning critical 75 inclusive, fatal 100 inclusive + 9999 included. The metric is measured every so often and more often (99%) of the time it's normal. I want to visually represent that fact, together with the percentage of time that he spends in the strips of quick, critical and fatal alert threshold.
For the dashboard but mainly reports I am looking for the percentage of time that the metric through each of the bands of threshold and put them in some form of chart, preferably very similar to how SCOM it: -.
I would also like to increase this visually with a quantification for the oriented numercally, in order to insert values in the report for clarification would be great too, for example:-Normal:-99% 0.5% warning critical Fatal 0.1% 0.4%
I think of what I have already learned that we have to include a 'blue' band for threshold indefinite in order to operate on a regular basis for any measure.
I do not seem to come up with this concept in Foglight but I think it could be very useful to have something. Any help is most appreciated.
Health and alarms is a standard display which can be used on any object topology. You can access it from the data browser and should also be able to specify this dashboard as a preference in personalized dashboards.
Here it is in a custom dashboard:
-
Need help with Group functions
I'm a total novice with SQL, so please forgive me if the answer to my question seems to be too obvious
I work with diagrams of the sample (in particular with the employees table):
DESC employees;
result
What I have to do is select all the managers, including the number of subordinates is higher than the average number of subordinates of managers who work in the same Department. What I've done so far is as follows:
SELECT mgr.employee_id manager_id, Director of mgr.last_name, mgr.department_id, COUNT (emp.employee_id)
Employees emp employees JOIN Bishop
ON emp.manager_id = mgr.employee_id
GROUP OF mgr.employee_id, mgr.last_name, mgr.department_id
ORDER BY mgr.department_id;
result
As you can see, I'm almost done. Now, I need only to calculate the average of the result of the COUNT function for each Department. But I'm totally stuck at this point.
All advice?Hello
Welcome to the forum!
user12107811 wrote:
I'm a total novice with SQL, so please forgive me if the answer to my question seems to be too obviousJust the opposite! Looks like a very difficult mission.
I work with diagrams of the sample (in particular with the employees table):
DESC employees;
resultWhat I have to do is select all the managers, including the number of subordinates is higher than the average number of subordinates of managers who work in the same Department. What I've done so far is as follows:
SELECT mgr.employee_id manager_id, Director of mgr.last_name, mgr.department_id, COUNT (emp.employee_id)
Employees emp employees JOIN Bishop
ON emp.manager_id = mgr.employee_id
GROUP OF mgr.employee_id, mgr.last_name, mgr.department_id
ORDER BY mgr.department_id;
resultAs you can see, I'm almost done. Now, I need only to calculate the average of the result of the COUNT function for each Department. But I'm totally stuck at this point.
All advice?Yes, you're almost done. You just need to add one more condition. You have to calculate the average value of total_cnt (the COUNT (*) you already do) of a Department and compare that to total_cnt.
There are several ways to do this, including
a scalar subquery (in a HAVING clause)
(b) make a result set with one line per Department, containing the average_cnt and reach than your current result set
(c) analytical functions. Analytical functions are calculated after the GROUP BY clause is applied and aggregate functions are calculated, it is legitimate to say "AVG (COUNT (*)) MORE (...)").If thinking (c) is the simplest. It involves the use of a query of Tahina, but (a) and (b) also require subqueries.
This sounds like homework, so I'll do it for you.
Instead, here is a very similar problem with the hr.employees table.
Let's say that we are interested in total wages given each type of work in each Department.SELECT department_id , job_id , SUM (salary) AS sum_sal FROM hr.employees GROUP BY department_id , job_id ORDER BY department_id , job_id ;Results:
DEPARTMENT_ID JOB_ID SUM_SAL ------------- ---------- ---------- 10 AD_ASST 4400 20 MK_MAN 13000 20 MK_REP 6000 30 PU_CLERK 13900 30 PU_MAN 11000 40 HR_REP 6500 50 SH_CLERK 64300 50 ST_CLERK 55700 50 ST_MAN 36400 60 IT_PROG 28800 70 PR_REP 10000 80 SA_MAN 61000 80 SA_REP 243500 90 AD_PRES 24000 90 AD_VP 34000 100 FI_ACCOUNT 39600 100 FI_MGR 12000 110 AC_ACCOUNT 8300 110 AC_MGR 12000 SA_REP 7000Now suppose we want to find out which of these sum_sals is higher than the average sum_sal of his Department.
For example, in detriment 110 (near the end OIF the list) there two types of work (AC_ACCOUND and AC_MGR) that have sum_sals of 8300 and 12000. The average of these two numbers is 10150, so we selected AC_MGR (because its sum_sal, 12000, is superior to 10150, and we do not want to include AC_ACCOUNT, because its sum_sal, 8300, is less than or equal to the average of the Department.
In departments where there is only one job type (for example, Department 70, or null "Department" at the end of the list above) the only sum_sal will be the average; and because the sum_sal is not greater than the average, we want to exclude this line.Let's start with the calculation of the avg_sum_sal using the analytical function AVG:
SELECT department_id , job_id , SUM (salary) AS sum_sal , AVG (SUM (salary)) OVER (PARTITION BY department_id) AS avg_sum_sal FROM hr.employees GROUP BY department_id , job_id ORDER BY department_id , job_id ;Output:
DEPARTMENT_ID JOB_ID SUM_SAL AVG_SUM_SAL ------------- ---------- ---------- ----------- 10 AD_ASST 4400 4400 20 MK_MAN 13000 9500 20 MK_REP 6000 9500 30 PU_CLERK 13900 12450 30 PU_MAN 11000 12450 40 HR_REP 6500 6500 50 SH_CLERK 64300 52133.3333 50 ST_CLERK 55700 52133.3333 50 ST_MAN 36400 52133.3333 60 IT_PROG 28800 28800 70 PR_REP 10000 10000 80 SA_MAN 61000 152250 80 SA_REP 243500 152250 90 AD_PRES 24000 29000 90 AD_VP 34000 29000 100 FI_ACCOUNT 39600 25800 100 FI_MGR 12000 25800 110 AC_ACCOUNT 8300 10150 110 AC_MGR 12000 10150 SA_REP 7000 7000Now all we have to do is to compare the sum_sal and avg_sum_sal columns.
Given that the analytic functions are calculated after the WHERE clause is applied, we cannot use avg_sum_sal in the WHERE clause of the query, even where it has been calculated. But we can do that in a subquery; Then, we can use avg_sum_sal in any way that we love in the Super-requete:WITH got_avg_sum_sal AS ( SELECT department_id , job_id , SUM (salary) AS sum_sal , AVG (SUM (salary)) OVER (PARTITION BY department_id) AS avg_sum_sal FROM hr.employees GROUP BY department_id , job_id ) SELECT department_id , job_id , sum_sal FROM got_avg_sum_sal WHERE sum_sal > avg_sum_sal ORDER BY department_id , job_id ;Results:
DEPARTMENT_ID JOB_ID SUM_SAL ------------- ---------- ---------- 20 MK_MAN 13000 30 PU_CLERK 13900 50 SH_CLERK 64300 50 ST_CLERK 55700 80 SA_REP 243500 90 AD_VP 34000 100 FI_ACCOUNT 39600 110 AC_MGR 12000 -
More help with analytical functions
I had great hellp here yesterday and I need once more today. I guess I'm still not able to get a solid understanding of analytical functions. So here's the problem:
table with 3 collars:
product_id (int), sale_date (to date), count_sold (int) - each file show that the number of items have been sold for the product at a given date.
The query should return the 3 passes of the table AND a fourth column that contains the date with the best sales of the product. If there are two or more dates with equal sales, the last being is chosen.
Is this possible using an analytical function appropriately and without using a subquery?
example:
product_id, sale_date, count_sold, high_sales_date
1, 01-01-2008, 10, 05/10/2008,.
1, 2008-03-10, 20, 10/05/2008
1, 10/04/2008, 25, 05/10/2008
1, 10/05/2008, 25, 05/10/2008
1, 01/06/2008, 22, 05/10/2008
2, 05/12/2008, 12, 05/12/2008
2, 06/01/2009, 10, 05/12/2008
Thank youHello
Try this:
SELECT product_id , sale_date , count_sold , FIRST_VALUE (sale_date) OVER ( PARTITION BY product_id ORDER BY count_sold DESC , sale_date DESC ) AS high_sales_date FROM table_x;If you would post INSERT statements for your data, then I could test it.
Focus issue: Why use FIRST_VALUE with descending order and not LAST_VALUE (ASCending) ORDER of default?
-
First off the coast to let put me the warning that I am not the right Wick informed in PL/SQL. I use this for an example posted in the forum of the APEX.
I need the function to return values based on the user role. I know, it's probably the ugliest pl/sql, you may have seen yet ;) I hope that you guys can decipher the logic that I use. Obviously, this code does not validate.
You guys could help me with this?
FUNCTION to CREATE or REPLACE return_art_lov_fn
RETURN art_table_type
AS
v_data art_table_type: = art_table_type ();
BEGIN
IF
EXISTS (SELECT ' 1' FROM GBL_PEOPLE, GBL_ACCESS WHERE upper (gbl_people.userid) = upper (app_user) AND gbl_people.person_id = gbl_access.person_id and gbl_access.art_role = 9)
THEN
FOR c IN (SELECT reverse_name, person_id
OF gbl_people)
LOOP
v_data. EXTEND;
v_data (v_data. (COUNTY): = art_rectype (c.person_id, c.reverse_name);
END LOOP;
ON THE OTHER
FOR c IN (select reverse_name, person_id of GBL_PEOPLE where upper (userid) = upper (APP_USER) and current_flag is not null
Union
Select reverse_name, person_id from GBL_PEOPLE where mgr_person_id = (select person_id in GBL_PEOPLE where upper (userid) = upper (APP_USER) and)
(current_flag is not null) and current_flag is not null)
LOOP
v_data. EXTEND;
v_data (v_data. (COUNTY): = art_rectype (c.person_id, c.reverse_name);
END LOOP;
ENDIF;
RETURN v_data;
END;
-VinodMy guess is that you want something like
CREATE OR REPLACE FUNCTION return_art_lov_fn RETURN art_table_type AS v_data art_table_type; l_cnt integer; BEGIN SELECT 1 INTO l_cnt FROM dual WHERE EXISTS( SELECT 1 FROM gbl_people p, gbl_access a WHERE upper( p.userid ) = upper( app_user ) AND p.person_id = a.person_id AND a.art_role = 9 ); IF( l_cnt = 1 ) THEN SELECT art_rectype( person_id, reverse_name ) BULK COLLECT INTO v_data FROM gbl_people; ELSE SELECT art_rectype( person_id, reverse_name ) BULK COLLECT INTO v_data FROM (SELECT person_id, reverse_name FROM gbl_people WHERE upper(userid) = upper(app_user) AND current_flag is not null UNION SELECT person_id, reverse_name FROM gbl_people WHERE mgr_person_id = (SELECT person_id FROM gbl_people WHERE upper(userid) = upper( app_user ) AND current_flag is not null) AND current_flag is not null); END IF; RETURN v_data; END;If this does not work (I do not have definitions table, so I can't try to compile it myself), please report all errors of compilation (including line numbers).
Justin
-
Help with analytical functions
Hi all
I'm on Oracle 11g DB and have records in the table that look like this
Analytical, I generate rownumber by Ref single transaction as follows:transaction_ref line_type description -------------------- -------------- --------------- 10 DETAIL abc123 10 DETAIL abc978 10 DETAIL test 10 DETAIL test 10 DETAIL test 20 DETAIL abcy 20 DETAIL abc9782 20 DETAIL test12 20 DETAIL test32
However, for my needs, I need my rownumber as follows:SELECT row_number() over (partition by transaction_ref order by 1) rownumber FROM mytable ; transaction_ref line_type description rownumber -------------------- -------------- --------------- ---------------- 10 DETAIL abc123 1 10 DETAIL abc978 2 10 DETAIL test 3 10 DETAIL test 4 10 DETAIL test 5 20 DETAIL abcy 1 20 DETAIL abc9782 2 20 DETAIL test12 3 20 DETAIL test32 4
with the exception of number 1 of Clotilde, I want to increment the number of lines per 3
Thank youtransaction_ref line_type description rownumber -------------------- -------------- --------------- ---------------- 10 DETAIL abc123 1 10 DETAIL abc978 4 10 DETAIL test 7 10 DETAIL test 10 10 DETAIL test 13 20 DETAIL abcy 1 20 DETAIL abc9782 4 20 DETAIL test12 7 20 DETAIL test32 10 ....
Maëlle
Published by: user565538 on June 4, 2011 17:32
Published by: user565538 on June 4, 2011 17:34
Published by: user565538 on June 4, 2011 17:35with mytable as ( select 10 transaction_ref,'DETAIL' line_type,'abc123' description from dual union all select 10,'DETAIL','abc978' from dual union all select 10,'DETAIL','test' from dual union all select 10,'DETAIL','test' from dual union all select 10,'DETAIL','test' from dual union all select 20,'DETAIL','abcy' from dual union all select 20,'DETAIL','abc9782' from dual union all select 20,'DETAIL','test12' from dual union all select 20,'DETAIL','test32' from dual ) SELECT transaction_ref, line_type, description, (row_number() over (partition by transaction_ref order by 1) - 1) * 3 + 1 rownumber FROM mytable / TRANSACTION_REF LINE_T DESCRIP ROWNUMBER --------------- ------ ------- ---------- 10 DETAIL abc123 1 10 DETAIL abc978 4 10 DETAIL test 7 10 DETAIL test 10 10 DETAIL test 13 20 DETAIL abcy 1 20 DETAIL abc9782 4 20 DETAIL test12 7 20 DETAIL test32 10 9 rows selected. SQL>SY.
-
Oracle 9i
The tables are
The function is respectivelyCREATE TABLE VEHICLE_CLASS(VEHICLE_CLASS CHAR(2),CAR_LD VARCHAR2(50)); INSERT INTO vehicle_class VALUES('K','130 STANDARD'); CREATE TABLE VEHICLE_BODY(V_BODY CHAR(2),CAR_LD VARCHAR2(50)); INSERT INTO vehicle_body VALUES('B','3 DOOR STATION WAGON');
ENTER PASTCREATE FUNCTION V_CLASS(V_IN VARCHAR2) RETURN VARCHAR2 AS veh_type VARCHAR2(20):=substr(v_in,1,2); veh_class VARCHAR2(20):=substr(v_in,3,1); veh_body VARCHAR2(20):=SUBSTR(v_in,4,1); veh_rtn varchar2(50); vehicle_found char; VEHICLE_BODY_FOUND CHAR; BEGIN IF veh_type='LD' THEN DBMS_OUTPUT.PUT_LINE('FOUND'); SELECT 'y' INTO vehicle_found FROM vehicle_class WHERE v_class=veh_class AND defender_ld_desc IS NOT NULL; DBMS_OUTPUT.PUT_LINE('FOUND2'); SELECT 'y' INTO VEHICLE_BODY_FOUND FROM vehicle_body WHERE v_body=veh_body AND defender_ld_desc IS NOT NULL; DBMS_OUTPUT.PUT_LINE('FOUND 3'); IF vehicle_found='y' THEN IF VEHICLE_BODY_FOUND='y' THEN veh_rtn := 'VALID'; ELSE veh_rtn := 'CAR - INVALID BODY STYLE CODE'; END IF; ELSE veh_rtn := 'CAR - INVALID CLASS CODE'; END IF; END IF; RETURN veh_rtn; EXCEPTION WHEN NO_DATA_FOUND THEN RETURN 'INVALID'; WHEN OTHERS THEN RETURN 'OTHER PROBLEM'; END V_CLASS;
'LDKB '.
THE OUTPUT IS
Kindly help me as to why the function returns invalid, it is not in the select statement, even if the value is valid.LDKB FOUND INVALIDThe problem is in your definition of the table...
CREATE TABLE VEHICLE_CLASS(VEHICLE_CLASS CHAR(2),CAR_LD VARCHAR2(50)); INSERT INTO vehicle_class VALUES('K','130 STANDARD'); CREATE TABLE VEHICLE_BODY(V_BODY CHAR(2),CAR_LD VARCHAR2(50)); INSERT INTO vehicle_body VALUES('B','3 DOOR STATION WAGON');Your vehicle_class and v_body are defined as CHAR (2) so when you insert the value 'B' and 'K' in those he's filling them with spaces to 2 characters.
Best bet is to avoid using TANK in your table definitions, unless you use only single character flags (for example, CHAR (1)) in which case it will not have adverse and simply stick to use instead of VARCHAR2.
-
I have a HP G60 554 CA most of the time of love - need help with the functions of the keyboard.
On my HP G60 554 CA, I do a lot of accounting/bookkeeping. Suddenly, I have no idea what I did but my alphabet keyboard turns suddenly to the BLUE character on the key of the alphabet. For example - the question mark & / key (white color) is equivalent to using the blue French E and e I know it's to do with the FN or fn key but I don't know how to tell Bank to regular. Any help would be appreciated.
Thank you.
Take a look at this link
-
Help with analytical functions - Windowing
Hello
I'm using Oracle 11.2.0.4.0.
I want to do the sum of all amounts for each window of 3 days from the date of the oldest rolling. I also want to name each window with the date limit for the period of 3 days.
My requirement is slightly more complicated, but I use this example to illustrate what I'm trying to
create table test (dt date, amt, run_id number);
Insert test values (to_date (' 22/04/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 23/04/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 24/04/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 25/04/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 27/04/2015 ',' dd/mm/yyyy'), 5, 1);
Insert test values (to_date (' 28/04/2015 ',' dd/mm/yyyy'), 2, 1);
Insert test values (to_date (' 29/04/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 04/30/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 01/05/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 02/05/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 03/05/2015 ',' dd/mm/yyyy'), 1, 1);
Insert test values (to_date (' 04/05/2015 ',' dd/mm/yyyy'), 1, 1);
The output should look like the example below. The period column requires
to show the end of each 3-day study:
AMT DT SUM_PER_PERIOD PERIOD
22/04/2015 1 1 24/04/2015
23/04/2015 1 2 24/04/2015
24/04/2015 1 3 24/04/2015
25/04/2015 1 3 27/04/2015
27/04/2015 5 6 27/04/2015
28/04/2015 2 7 30/04/2015
29/04/2015 20 27 30/04/2015
30/04/2015 30 52 30/04/2015
05/01/2015 5 55 3/05/2015
05/02/2015 5 50 3/05/2015
05/02/2015 10 50 3/05/2015
05/03/2015 1 21/3/05/2015
All I can manage this is
Select dt
TN
, sum (amt) on sum_per_period (PARTITION BY run_id ORDER BY dt vary from 2 PAST current line)
of the test
order by dt;
Can anyone help?
It's very kind of you to give the insert and create instructions... but I corrected the data a bit
It does not match the output see you below
starting from 29/04, you forgot to change the dates and numbers of...
insert into test values (to_date('22/04/2015','dd/mm/yyyy'),1,1); insert into test values (to_date('23/04/2015','dd/mm/yyyy'),1,1); insert into test values (to_date('24/04/2015','dd/mm/yyyy'),1,1); insert into test values (to_date('25/04/2015','dd/mm/yyyy'),1,1); insert into test values (to_date('27/04/2015','dd/mm/yyyy'),5,1); insert into test values (to_date('28/04/2015','dd/mm/yyyy'),2,1); insert into test values (to_date('29/04/2015','dd/mm/yyyy'),20,1); insert into test values (to_date('30/04/2015','dd/mm/yyyy'),30,1); insert into test values (to_date('01/05/2015','dd/mm/yyyy'),5,1); insert into test values (to_date('02/05/2015','dd/mm/yyyy'),5,1); insert into test values (to_date('02/05/2015','dd/mm/yyyy'),10,1); insert into test values (to_date('03/05/2015','dd/mm/yyyy'),1,1);your periods will change if you insert a new first date...
so I guess you want a specific date... in this case 22/04/2015 and a specific end date
creation of periods from this first date and then grouping of these periods is easier with a first fixed date and a delta of 3 days.
the first step is to match the periods to your data (adapted)
with periods as ( select date_start + (level-1) * period_days period_start, date_start + level * period_days period_end, period_days from ( select to_date('21/04/2015', 'dd/mm/yyyy') date_start, to_date('04/05/2015', 'dd/mm/yyyy') date_end, 3 period_days from dual) connect by date_start + level * period_days < date_end) select * from test t, periods p where t.dt > p.period_start and t.dt <= p.period_endThis gives your data with the dates of beginning and ending period
DT AMT RUN_ID PERIOD_START PERIOD_END PERIOD_DAYS 22/04/2015 1
1
21/04/2015 24/04/2015 3
23/04/2015 1
1
21/04/2015 24/04/2015 3
24/04/2015 1
1
21/04/2015 24/04/2015 3
25/04/2015 1
1
24/04/2015 27/04/2015 3
27/04/2015 5
1
24/04/2015 27/04/2015 3
28/04/2015 2
1
27/04/2015 30/04/2015 3
29/04/2015 20
1
27/04/2015 30/04/2015 3
30/04/2015 30
1
27/04/2015 30/04/2015 3
05/01/2015 5
1
30/04/2015 05/03/2015 3
05/02/2015 5
1
30/04/2015 05/03/2015 3
05/02/2015 10
1
30/04/2015 05/03/2015 3
05/03/2015 1
1
30/04/2015 05/03/2015 3
and then sum the amt during the 3 days
with periods as ( select date_start + (level-1) * period_days period_start, date_start + level * period_days period_end, period_days from ( select to_date('21/04/2015', 'dd/mm/yyyy') date_start, to_date('04/05/2015', 'dd/mm/yyyy') date_end, 3 period_days from dual) connect by date_start + level * period_days < date_end) select t.dt, t.amt, sum(amt) over (order by t.dt range between 2 preceding and current row) sum_per_period, p.period_end period from test t, periods p where t.dt > p.period_start and t.dt <= p.period_endgiving your output as requested:
DT AMT SUM_PER_PERIOD PERIOD 22/04/2015 1
1
24/04/2015 23/04/2015 1
2
24/04/2015 24/04/2015 1
3
24/04/2015 25/04/2015 1
3
27/04/2015 27/04/2015 5
6
27/04/2015 28/04/2015 2
7
30/04/2015 29/04/2015 20
27
30/04/2015 30/04/2015 30
52
30/04/2015 05/01/2015 5
55
05/03/2015 05/02/2015 5
50
05/03/2015 05/02/2015 10
50
05/03/2015 05/03/2015 1
21
05/03/2015 -
Need help with the function MAX Returns the values
I am creating a report to return slow movement inventory data. One of the requests is that it return only the last date that an item is traded on. A map shows the last date of receipt for a part, the other will show the last time, a part has been issued or shipped on a sales order.
The hiccup is that it returns every single of the last time that an item has been received, so every last show of the material (on the second sheet) of items on-site.
Could someone help me set the max value? As shown below, and many variations, the sheet comes up with no corrupt data or dates.
MAX (Date of the MAX operation) (PARTITION OF matter Transactions.Item matter Transactions.Item ORDER)
Always returns the next two when in reality, I want just the one with the date the most recent (April 2010).
100034 BNDSCE-105 - QUALITY BEARINGS OR EQUIVALENT A400M AB01D... Problem of component WIP $0.00 11-Sep-2009-3
100034 BNDSCE-105 - QUALITY BEARINGS OR EQUIVALENT A400M AD01D... Problem of component WIP $0.00 April 13, 2010-16
Thank you for your help.
BeckaHi Becka
If you have a new calculated column showng, as in your example, 13 April 10, and you have another column that contains also this date, you should be able to create a new condition. Condition that must be the date calculated max is equal to the actual date, thus returning only the line that includes 13 April 10.I hope I'm makng sense
Best wishes
Michael -
Need help with to_char function
Hi all
I use sql query below to get sysdate with timestamp.
Select to_char (' sysdate, ' YYYYMMDD HH24:MM:SS ") of the double
It gives me below the result: 20130816 05:08:49
Strangely, when I keep on run the same query, I noticed after 60 seconds are completed, timestamp begins with 20130816 05:08:01 again.
I'm missing something here...
Thanks in advance
998158 wrote:
Select to_char (' sysdate, ' YYYYMMDD HH24:MM:SS ") of the double
Strangely, when I keep on run the same query, I noticed after 60 seconds are completed, timestamp begins with 20130816 05:08:01 again.
Hello
You use Mi minutes not MM.
Use like this
Select to_char (sysdate, "YYYY/MM/DD HH24:Mi:SS) twice;
See you soon!
-
Simple problem: need help with string functions
I'm sure it's easy, but it's late and I'm tired.
If I had a string that contained a full name for example.
< cfset name = 'John Smith' >
How would manipulate the string so I could separate in two variables for example.
FirstName = 'John '.
LastName = 'Smith '.
See you soonSwampie says:
> How would manipulate the string so I could separate in two variables for example.
> firstname = 'John '.
> lastname = 'Smith '.pretend it's a list delimited by spaces.
name = 'Jane Doe';
Switch (listLen (name,"" ")) {}
case 1:
manage single names
break;
case 2:
firstName = listFirst (name,"" ");
lastName = listLast (name,"" ");
break;
case 3:
firstName = listFirst (name,"" ");
middleName = listGetAt ("name, 2," "");
lastName = listLast (name,"" ");
break;
} / / switchprobably need a break by default for > 3 names.
-
version 9.2
Here is a sample
Output: number of separate orders for each dateWITH temp AS (SELECT 10 ID, TRUNC (SYSDATE - 1) dt, 101 ord_id FROM DUAL UNION SELECT 11 ID, TRUNC (SYSDATE - 1) dt, 101 ord_id FROM DUAL UNION SELECT 11 ID, TRUNC (SYSDATE) dt, 103 ord_id FROM DUAL UNION SELECT 13 ID, TRUNC (SYSDATE) dt, 104 ord_id FROM DUAL) SELECT * FROM temp
Dt Count 1/25 1 1/26 2ME_XE?WITH temp AS 2 (SELECT 10 ID, TRUNC (SYSDATE - 1) dt, 101 ord_id 3 FROM DUAL 4 UNION 5 SELECT 11 ID, TRUNC (SYSDATE - 1) dt, 101 ord_id 6 FROM DUAL 7 UNION 8 SELECT 11 ID, TRUNC (SYSDATE) dt, 103 ord_id 9 FROM DUAL 10 UNION 11 SELECT 13 ID, TRUNC (SYSDATE) dt, 104 ord_id 12 FROM DUAL) 13 SELECT dt, count(distinct ord_id) 14 FROM temp 15 group by dt; DT COUNT(DISTINCTORD_ID)-------------------------- ---------------------25-JAN-2009 12 00:00 126-JAN-2009 12 00:00 2 2 rows selected. Elapsed: 00:00:00.01ME_XE?ME_XE?
-
Hello
I have a data in table (raj_table) with columns (char11) raj_id, raj_number (varchar2 (15)), raj_format (NUMBER), Primary_ID (identity with the values of the primary key column)
Primary_ID raj_id Raj_number Raj_format
1 raj rajvend 1
2 raj rajvend 1
3 raj rajvendor1 2
4 raj rajvendor1 2
5 raj rajvendor1 2
6 raj rajvendor2 3
I used under SQL to get query output as below, but has not achieved the required result:
Select client_id vendor_number, vendor_format, primary_id, row_number() on sl_no (client_id partition, primary_id, vendor_format order of client_id primary_id, vendor_format, vendor_number, vendor_number)
from raj_table by sl_no asc
SL_NO raj_id raj_number raj_format primary_id
1 1 raj rajvendor 1
1 2 raj rajvendor 1
2 3 raj rajvendor1 2
2 4 raj rajvendor1 2
2 5 raj rajvendor1 2
3 6 raj rajvendor2 3
I need help with a SQL query to get the result as above without using the group by clause. I want to bring together the combination of separate line of the three columns (raj_id, raj_number, raj_format) and add a unique serial number for each online game (SL_NO column below). So, above there are 3 unique set of (raj_id, raj_number, raj_format) I can get in a group by clause, but I can not add prmiary_id, SL_NO values if I group by clause. I used the analytical functions like row_number() but no luck. Need solution for this.
with t as)
Select 'raj' raj_id, 'rajvend' raj_number, 1 raj_format, 1 primary_id Union double all the
Select option 2, 'raj', 'rajvend', 1 double Union all
Select 3, 'raj', 'rajvendor1', 2 double Union all
Select 4, 'raj', 'rajvendor1', 2 double Union all
Select 5, 'raj', 'rajvendor1', 2 double Union all
Select 6, 'raj', 'rajvendor2', 3 double
)
Select dense_rank() over (order of raj_id, raj_number, raj_format) sl_no,
t.*
t
order by primary_id
/
PRIMARY_ID RAJ RAJ_NUMBER RAJ_FORMAT SL_NO
---------- ---------- --- ---------- ----------
1 1 raj rajvend 1
1 2 raj rajvend 1
2 3 raj rajvendor1 2
2 4 raj rajvendor1 2
2 5 raj rajvendor1 2
3 6 raj rajvendor2 36 selected lines.
SQL >
SY.
-
I'm suddenly needing help with my browser Firefox (6.0.2)
Hello
I'm suddenly needing help with my browser Firefox (6.0.2)(OS: I use Windows XP).
When I open the browser, I don't see is a totally white screen of white, with all the toolbars at the top.
I know that my physical connections are very good: I have tested the modem, turned the pc market etc and I can also receive/send emails.
This problem started today, September 8, 2011 and has never happened before.
Is it a coincidence that Firefox itself to day before I disconnected yesterday evening? Could this be something to do with this particular new update?
I also noticed that just before I "opened" Firefox, I now get a small box indicating:
[JAVASCRIPT APPLICATION]
Handl exc in Ev: TypeError: this oRoot.enable is not a functionThis never appeared before - I hope that it offers a clue has what is wrong.
The browser not be stuck in Mode safe, said by the way.
Of course, I can't find solutions to the problem on the internet, I don't physically see all Web sites!
(A friend sends this request in my name from their pc)Any light you can throw on this problem of confusion would be much appreciated. I'd rather not have to uninstall and reinstall Firefox if possible.
If the only option is to uninstall Firefox and reinstall from your site, I'm also in trouble (I can not see the internet or download).
In this case, would you be able to send the .exe file as an attachment to my e-mail address? In the affirmative, please let me know and I'll give you more details.Thanks in advance.
One possible cause is security software (firewall) that blocks or limits Firefox or plugin-container process without informing you, possibly after the detection of changes (update) for the Firefox program.
Delete all rules for Firefox in the list of permissions in the firewall and leave your firewall again ask permission to get full unlimited access to the internet for Firefox and the plugin-container and the update process.
See:
Start Firefox in Firefox to solve the issues in Safe Mode to check if one of the extensions of the origin of the problem (switch to the DEFAULT theme: Firefox (Tools) > Add-ons > appearance/themes).
- Makes no changes on the start safe mode window.
- https://support.Mozilla.com/kb/safe+mode
Maybe you are looking for
-
The MIG 229771 in c# software error
Hello I use an NI USB-6009 data acquisition. I'm unable to create a task using c#, I get a MIG 229771 software error when I try to instantiate with 'myTask1 task = new Task();' or 'task myTask1 = DaqSystem.Local.LoadTask ("D00");' but not when I crea
-
Hi guys,.my video plays really lag then I play my laptop, but if I use it for my ipad is not... so it means there is sumthing wrong with my browsers or sumthing. can you guys tell me the solution for this please?
-
Currently, I'm trying to understand how to implement a network dropout detector in my application. I use HTTP and so often keep a socket opened for a long time and do a lot of things on it. Therefore, a random network dropping out really hurts to e
-
Use of automatic start and RuntimeStore for UIApplication... Help!
I have a UIApplication that runs in the background and uses another entry point. The problem I have is to initialize the high device. The application will be auto-start and being thrust into the RuntimeStore (it seems anyway), and it will run in the
-
BlackBerry Smartphones shipper receiver postmaster message - not delivered on my handheld
Hello, I recently canceled my blackberry phone, and I'm syncing my @me.com email with mobile e-mail account @mobileemail.vodafone.com.au. All of you who sent me an email to my account @me.com receive an email back saying that I him have not received