SUM TOP N Values Using Rank?
Jul 20, 2010
I certainly know this is possibly but I am trying to do this on the fly and can't seem to work it out:
I have a table A:
ID Name Priority
-------------------
1 Smith 1
1 SSmith 2
1 ASmith 3
1 BSmith 3
2 John 2
3 Ed 1
and I am looking to create the following table from this:
ID Name Sum(Top3Priority)
--------------------------------
1 Smith,SSmith,etc 8
Now, I've got listagg working and everything appears to be going swimmingly but: for every listagg grouping on name I need to only sum the highest top 3 priorities. So in the example above there are four Smiths but I need to only sum the top 3 priorities which are 3,3,2 and ignore the 1 even though I do want all the listagg Smith's (SSmiht, ASmith, etc) in there.
Now I can sum the priority, but don't really know how to sum only the top 3 in any ID ? There can be 1 to n ID's so if there are only 2 ID's I want to sum those 2, if there are 3 all 3 and 4 upwards only the top 3.Here is a snippet of the SQL I am using
SLECT id,
listagg(MN_CR_LOOKUP.f_name, ',') within group (order by Priority)) roadname,
**** sum top 3 here ?****
count(*) "NumI", Sum("Elevation") "CombinedElevation"
FROM jc,
[code]...
View 3 Replies
ADVERTISEMENT
Mar 24, 2013
I created a sample table named as "Student" with following data. table contains two columns only - stdid & marks.
stdid marks
10 75
20 60
30 60
40 45
50 30
I have to find the rank of students based on their marks in descending order.Is there a way to get rank without using RANK function?
View 11 Replies
View Related
Mar 10, 2013
I am trying to find out the difference in time between peoples memberships and also the order that these memberships are taken out in. So far I have added in a rank statement to work out the order the memberships were created in, but now want to look at the difference between the dates returned. The SQL I used is:
SELECT owner_party_id,
mem_number,
support_id,
mem_start_date,
RANK() OVER (PARTITION BY owner_party_id ORDER BY mem_start_date ASC) MEMBERSHIP_SEQUENCE
FROM membership_all
WHERE version_type = 'CUR'
AND owner_party_id IN ('65051', '65051', '65348', '65348', '65607', '65607', '65607')
to get:
"OWNER_PARTY_ID"|"MEM_NUMBER"|"SUPPORT_ID"|"MEM_START_DATE"|"MEMBERSHIP_SEQUENCE"
65051|318874751|8014747|01-MAR-10|1
65051|412311060|21502883|15-AUG-12|2
65348|308672459|3526913|01-MAY-10|1
65348|409951130|20950524|18-JUN-12|2
65607|315830192|7510133|17-MAY-10|1
65607|406448110|20024246|16-MAR-12|2
65607|409738130|20903556|14-JUN-12|3
Now I would like to calculate the difference between the start dates of each of the owner_party_id groups, so to get something like this:
OWNER_PARTY_ID|MEM_NUMBER |SUPPORT_ID|MEM_START_DATE |MEMBERSHIP_SEQUENCE|Diff
65051|318874751|8014747|01-Mar-10|1|
65051|412311060|21502883|15-Aug-12|2|898
65348|308672459|3526913|01-May-10|1
65348|409951130|20950524|18-Jun-12|2|779
65607|315830192|7510133|17-May-10|1
65607|406448110|20024246|16-Mar-12|2|669
65607|409738130|20903556|14-Jun-12|3|90
I think that I need to use the Lag function in, but I am not too sure if it can be linked to look at the data within a grouping of owner party id, as it would make no sense to calculate the difference in dates for two different owner party ids.
View 4 Replies
View Related
Mar 27, 2013
I need to return which hour for a given date range had the most calls. I have a query that works but it is inelegant and I'm pretty sure I could do better. I'm pretty new to analytic queries so go easy...
select hour,
calls
from
(
select hour,
calls,
rank() over (ORDER BY calls desc) as ranking
from
(
[Code]....
View 1 Replies
View Related
Dec 2, 2010
how do i retrieve only last rank row from table ?
View 7 Replies
View Related
Oct 4, 2013
I've below table.
with tmp_tbl as
(select 'H1' as hh_id, 'C1' as cust_id, 2 as f_rnk, 'F' as gender, to_date('20130102','YYYYMMDD') as purch_dt, to_date('20100203','YYYYMMDD') first_dt from dual
union
select 'H1' as hh_id, 'C2' as cust_id, 1 as f_rnk, 'M' as gender, to_date('20130102','YYYYMMDD') as purch_dt, to_date('20100303','YYYYMMDD') first_dt from dual
union
select 'H1' as hh_id, 'C3' as cust_id, cast(null as number) as f_rnk, 'U' as gender, to_date('20130103','YYYYMMDD') as purch_dt,
[code].....
Now i need to rank each cust_id in each hh_id based on below conditions.
1) If atleaset one cust_id in hh_id has f_rnk then gender 'F' with highest f_rnk (more then one F with same f_rnk then the one with oldest first_dt), if no 'F' then gender 'U' with highest f_rnk ((more then one F with same f_rnk then the one with oldest first_dt)), if no 'F' and 'U' then consider 'M' (more then one M with same f_rnk then the one with oldest first_dt).
2) If the above is not met (no cust_id in hh_id has f_rnk populated) then i've to rank based on purch_dt. Gender 'F' with recent purch_dt (if more than one F in household with same purch_dt then the one with oldest first_dt), if no 'F' then gender 'U' with recent purch_dt (if more than one U in household with same purch_dt then one with oldest first_dt), if no 'F' and 'U' then consider 'M' (more than one M in household with same purch_dt then the one with oldest first_dt).
3) If the above criteria is also not met, then rank based on gender_cd. Gender 'F' will have first preference then 'U' and then 'M'.
My output :
HH_ID CUST_ID F_RNK GENDER PURCH_DT FIRST_DT F_RNK_RANK PURCH_RANK GENDER_ONLY_RANK
----- ------- ---------- ------ ----------- ----------- ------------ ------------ -----------------
H1 C1 2 F 1/2/2013 2/3/2010 1
H1 C2 1 M 1/2/2013 3/3/2010 2
H1 C3 U 1/3/2013 4/3/2010 3
H2 C4 F 4/3/2013 10/2/2009 2
H2 C5 M 5/5/2013 8/8/2010 1
H3 C6 F 5/6/2008 1
H3 C6 M 7/8/2010 2
I've tried below query with one condition, but it's giving f_rnk_rank for all records. How can i include multiple conditions in the rank function.
with tmp_tbl as
(select 'H1' as hh_id, 'C1' as cust_id, 2 as f_rnk, 'F' as gender, to_date('20130102','YYYYMMDD') as purch_dt, to_date('20100203','YYYYMMDD') first_dt from dual
union
select 'H1' as hh_id, 'C2' as cust_id, 1 as f_rnk, 'M' as gender, to_date('20130102','YYYYMMDD') as purch_dt, to_date('20100303','YYYYMMDD') first_dt from dual
union
select 'H1' as hh_id, 'C3' as cust_id, cast(null as number) as f_rnk, 'U' as gender, to_date('20130103','YYYYMMDD') as purch_dt, s hh_id, 'C5' as cust_id,
[code]....
View 3 Replies
View Related
Jul 13, 2011
there is a way when doing the Rank Function in PL/SQL to pass the field that will be ranked as an override.
SELECT rank(p_ColumnAmt) within group (order by p_ColumnNm desc) rank
INTO v_RnkNoAmt
FROM Table_name
WHERE ??????;
p_Column is the amount I am ranking
p_ColumnNm is the actual field name to Rank.
When I pass the field name with an override I do not get the correct rank back. If I run the Select with the actual field name curr_1_mth_amt, I get the correct rank.
I have about 70 different field to Rank and do not want to make a procedure for each field.
View 7 Replies
View Related
Dec 2, 2010
the following statement gives each customer owns how many promotions.
Q:) how do i retrieve a customer who has max. promotions?
SELECT C.CUSTOMER_ID,COUNT(P.PROMOTION_ID)
FROM PROMOTIONS P,CUSTOMERS C
WHERE C.CUSTOMER_ID = P.CUSTOMER_ID
GROUP BY C.CUSTOMER_ID
=====================================================
CUSTOMER_ID COUNT(P.PROMOTION_ID)
-------------------------------------
001 | 5
002 | 8
003 | 4
004 | 6
005 | 5
006 | 3
View 7 Replies
View Related
Feb 1, 2012
I have a SQL statement that return the following result.
col1
----
a
a
b
b
a
c
c
c
c
What i need is to eliminate the duplication and return the result as the following.
col1
----
a
b
a
c
Using group by or distinct will eliminate the third row "a" and will not return it ordered.My idea is to generate another column that rank the result as following.
col1rank
------------
a1
a1
b2
b2
a3
c4
c4
c4
c4
so using the following SQL query will return the result as needed
select col1, rank
from T
group by col1, rank
order by rank;
col1rank
------------
a1
b2
a3
c4
View 8 Replies
View Related
Mar 26, 2012
I have this code which simply gives me the count of term.
SELECT
OUA_ID
,COALESCE (MAX( DECODE( TERM, 0603, TERM_COUNT, NULL ) ), 0 ) "TERM_0603_CNT"
,COALESCE (MAX( DECODE( TERM, 0702, TERM_COUNT, NULL) ), 0 ) "TERM_0702_CNT"
,COALESCE (MAX( DECODE( TERM, 0705, TERM_COUNT, NULL ) ), 0 ) "TERM_0705_CNT"
,COALESCE (MAX( DECODE( TERM, 0707, TERM_COUNT, NULL ) ), 0 ) "TERM_0707_CNT"
FROM(
SELECT DISTINCT
OUA_ID, TERM_COUNT, TERM
FROM(
SELECT OUA_ID, TERM ,PROVIDER_CDE, COUNT(*) TERM_COUNT
FROM TABLE
WHERE PROVIDER_CDE = 'BILL'
GROUP BY OUA_ID, TERM, PROVIDER_CDE)) GROUP BY OUA_ID
Gives:
OUA_IDTERM_0603_CNTTERM_0702_CNTTERM_0705_CNTTERM_0707_CNT
32710 0 3 7 0
45726 2 0 1 0
Is there a way i could used the rank function (with decode) to only display the 1ST (First) and 2nd(Second) most used term per OUA_ID?
So i get this:
OUA_ID1st MOST USED2nd MOST USED
3271 705 702
45726 603 705
View 4 Replies
View Related
Aug 30, 2012
I have the following table :
CREATE TABLE A_TEST (A INTEGER, B INTEGER, C INTEGER, D INTEGER, FLAG CHAR(11));
INSERT INTO A_TEST (A,B,C,D) VALUES(1,2,3,4);
INSERT INTO A_TEST (A,B,C,D) VALUES(2,4,5,8);
INSERT INTO A_TEST (A,B,C,D) VALUES(1,2,3,4);
INSERT INTO A_TEST (A,B,C,D) VALUES(2,4,5,8);
[code].......
I would like to perform an update on the FLAG column by setting to "D" if it is a duplicate record.1,2,3,4);
I would like to use the rank function.
Desired update:
A B C D FLAG
1 2 3 4
2 4 5 8
1 2 3 4 D
2 4 5 8 D
7 2 3 4
9 2 3 4
7 2 3 4 D
1 2 3 4 D
5 4 5 8
2 2 3 9
2 4 5 8
6 2 3 4
1 3 3 4
8 2 8 4
View 5 Replies
View Related
Mar 17, 2013
can we create rank on a particular column without using rownum and rank function.
View 9 Replies
View Related
Jan 5, 2012
I am trying to use RANK() clause with a window clause...is that possible to use both together?
like
select col1, col2, col3,
RANK() OVER (ORDER BY col3 desc RANGE BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) RK
from table t
but getting error in SQL Developer
ORA-00907: missing right parenthesis
00907. 00000 - "missing right parenthesis"
*Cause:
*Action:
Error at Line: 2 Column: 33
The reason why i need to rank in window clause because i have data like this
Name Marks Quiz
Ali 10 1
John 20 1
Sara 30 1
John 40 2
Sara 50 2
Ali 20 2
...
...
and so on
I want to rank them based on their cumulative sum marks after every test..ranking should be in such a way that it should look current row and preceding rows
like this
Name Marks Quiz cumulative_marks rk
Ali 10 1 10 4
John 20 1 20 3
Sara 30 1 30 2
Peter 100 1 100 1
John 40 2 60 3 ==> becuase John has now third most overall cumulative marks (60) after quiz 2.
Sara 50 2 90 2 ==> becuase Sara has now 2nd most overall cumulative marks (90) after quiz 2.
Ali 20 2 30 4 ==> becuase Ali has now fouth most overall cumulative marks (30) after quiz 2.
View 11 Replies
View Related
May 1, 2013
Here table - tac has
row1
-----
X
X
X
B
[code]...
I want ouptut like using sql query with out using dense_rank function,
row1 row2
X 1
X 1
X 1
B 2
[code]...
View 7 Replies
View Related
Oct 22, 2012
I am attempting to user PIVOT and DENSE RANK in a query the following is the query and the record set it returns (condensed and de-identified)
"select * from(
select * from (select dense_rank() over (partition by 1 order by cal.weeksort desc) WEEK_nbr,
u.user_title Manager_Title, replace(hier.manager, '<br>',' - ') Manager,
replace(hier.user_hin, '<br>',' - ') user_name,
to_char(cal.calendar_date_week - 6, 'MM/DD/YYYY') ||' - '|| to_char(cal.calendar_date_week, 'MM/DD/YYYY') Week_of,
upper(substr(cal.day_of_week, 1,3)) DOW, count(distinct Pers_gen_key) cnt from apexim.hrw_member_action act
[code]....
View 2 Replies
View Related
Sep 1, 2010
Am working on a workbook to count the number of enrolments and withdrawals in the program. My data looks like this
name semester status year
A 1 enrol 2010
A 2 withdraw 2010
A 3 enrol 2010
B 1 enrol 2010
B 2 withdraw 2010
I want to count their latest status only. It should come up with
Total Enrol - 2
Total Withdrawn - 1
For total Withdrawn, I tried 'rank' and filter to equals 1 but it does not allow me. Is there any way to have this work?
Here's my calculation:(decode((FIRST_VALUE(status) OVER(PARTITION BY year, name ORDER BY semester DESC)),'withdraw', name)) It tells me that 'Aggregation of Analytic function not allowed'
View 3 Replies
View Related
Jul 25, 2013
I used Region, Process by to search the report which appears as shown above. Then I use Choose Auditors column to select my Auditor and copy paste it into the report under To be Audited By col. Is there a way to automate the process. I am here using a tabular form in APEX. My main aim is to assign auditors based on Region, not equal to Processed by.
View 4 Replies
View Related
Dec 3, 2010
I have a scenario where I have to get all the available dates of a resource. I am using the below query to get it.
Select Avail_Date AS MONTH
, Resource_Id
FROM res_tsk
, (SELECT Rownum - 1 + TRUNC (sysdate) avail_date
FROM Dual
[code].......
The result of this is:
Month Dates Resource_ID
12/3/10 0:00 NULL
12/4/10 0:00 NULL
12/5/10 0:00 NULL
12/6/10 0:00 100033868
As I am doing a outer join, if the resource is not available on a particular day the resource_id is coming as NULL as it is not available. Is there any way to populate this NULL resource_id with the original resource_id as the resource_id is same for all the result set.
I need the output to be
Month Dates Resource_ID
12/3/10 0:00 100033868
12/4/10 0:00 100033868
12/5/10 0:00 100033868
12/6/10 0:00 100033868
View 3 Replies
View Related
Dec 23, 2010
My scenario is to insert values into 'out' column by comparing 's' and 'IP' columns of temp table.The exact situation is at first need to go to ip column,take a value and then go to source column and check for the same value of ip which is taken previously.Then after corresponding ip of that source column should be inserted back in previous source column.
The situation is marked clearly in file which i am attaching with '--' comments at respective places.I am also pasting the code which i tried out,unfortunately it is giving error as exact fetch returns more than requested number of rows since there are duplicates in the table.I tried it using nested for loops.Also implemented using rowid,but it didnt work.
fixing the errors or if there is any new logic that can be implemented.
DECLARE
i_e NUMBER(10);
BEGIN
FOR cur_1 IN(SELECT IP from temp where IP IS NOT NULL)
LOOP
FOR cur_2 IN(SELECT IP from temp where s=cur_1.IP)
[Code]...
View 9 Replies
View Related
Oct 28, 2013
I am searching the simplest way for ad hoc MINUS.I do:
SELECT *
FROM uam_rss_user_XXXXXXX
WHERE host_name IN
('XXX0349',
'XXX0362',
'XXX0363',
'XXX0343',
'XXX0342',
'XXX0499',
[code]....
and look in the table which values are missing (values that are in host_name IN but not in actual table).is there a simpler way for doing an ad hoc MINUS? I know to insert values in temp. Table. How are experienced Oracle pros doing this task?
View 6 Replies
View Related
Jul 25, 2013
From two given tables, how do you fetch the values from two columns using values from one column(get values from col.A if col.A is not null and get values from col.B if col.A is null)?
View 2 Replies
View Related
Dec 18, 2012
I created a package with some types, and every was compiling fine. However, when I ran the new function, I got an error: ORA-21700: object does not exist or is marked for delete
After a little research, I realized that the types would have to be declared outside the package. As soon as I did that, I suddenly started getting the "not enough values" errors on all my types. I compared the number of columns being returned, and the number of columns in the type, and they match.Here is my type code:
CREATE OR REPLACE TYPE TSA_CUSTOM.Lost_Plan as object (
LP_Key number, -- The member key of the plan that is going away
LP_Type varchar2(20),
LP_Dept varchar2(12),
LP_SubDept varchar2(12),
LP_Class varchar2(12),
LP_VendorName varchar2(50)
[code]...
View 2 Replies
View Related
May 4, 2010
find the Test Case below.
--Creation of Table
create table tb1
(ID number(4),
event varchar2(20),
vdate date);
--Inserting Values into the Table.
INSERT ALL INTO tb1 (ID, event, vdate) VALUES (01, 'V1', '01-JAN-2009')
INTO tb1 (ID, event, vdate) VALUES (01, 'V2', '02-FEB-2009')
INTO tb1 (ID, event, vdate) VALUES (01, 'V3', '04-MAR-2009')
INTO tb1 (ID, event, vdate) VALUES (01, 'V4', '03-APR-2009')
INTO tb1 (ID, event, vdate) VALUES (01, 'V5', '05-MAY-2009')
[Code]...
--Selecting data from Table.
SELECT * FROM TB1;
ID EVENT VDATE
---------- -------------------- ---------
1 V1 01-JAN-09
1 V2 02-FEB-09
1 V3 04-MAR-09
1 V4 03-APR-09
1 V5 05-MAY-09
2 V1 01-JAN-10
2 V2 02-FEB-10
2 V3 04-MAR-10
2 V4 03-APR-10
2 V5 05-MAY-10
10 rows selected.
how can i display the data as below format using Oracle 9i SQL.
IDV1 V2 V3 V4 V5
--- ---------------- ------------ --------------- -------------- ------------
11-Jan-092-Feb-094-Mar-093-Apr-095-May-09
21-Jan-102-Feb-104-Mar-103-Apr-105-May-10
View 4 Replies
View Related
Jul 17, 2012
Suppose I have a table containing 2 columns and 5000 rows.
I need to take values that are at 259th, 561st, 872nd rows.(just example, these values are not fixed, can be anything from 1 to last row)
View 1 Replies
View Related
Mar 28, 2012
Here's what I need to do.
1) I need to find the max(peak_area) where the detecor = 'spectrum'.
2) Then I need to return peak_area for detectors (dad and elsd) of the highest corresponding peak_id.
sampledetectorpeak_idpeak_area
123dad175
123elsd180
123spectrum168
123dad290
123elsd260
123spectrum295
So in this very small dataset I need to return
123dad290
123elsd260
View 1 Replies
View Related
Sep 27, 2012
I have a table table1 with columns a,b,c,d
I am fetching some results from this table by means of this query:
SELECT
a,
b,
sum(case when c='F' then '1' else '0' end) "c_data",
sum(case when d='S' then '1' else '0' end) "d_data"
[Code]...
My result will be like this:
a b c_data d_data
abcd efg 1 1
asff fjd 1 0
dfd fd 0 0
Now i need a 5th column, which should look like this:-
a b c_data d_data Total
abcd efg 1 1 2
asff fjd 1 0 1
dfd fd 0 0 0
View 2 Replies
View Related
Jun 2, 2013
i want to add the values of rows in a table. for example if we enter marks for five subjects to a student we need to calculate the total marks and the average for that student. how to do this.
sample as
m1 m2 m3 m4 m5 total avg
== == == === === === ====
55 79 80 81 82 377 75.4
68 94 84 73 94 413 82.6
View 3 Replies
View Related
Nov 28, 2012
I have a dropdown in my jsp with all users from database. When I select a user I'm able to get the details of the corresponding result, but when I click on all it should give the total individually. Below is the code that I use to get individually.
<%@page contentType="text/html" pageEncoding="UTF-8"%>
<%@include file="DBCon.jsp" %>
<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta charset="utf-8" />
[code]...
View 1 Replies
View Related
Jul 23, 2010
im new to Database i have to count no of messages based on the minuties like 1-5, 6-15,16-30,>30 min how to write query
select count(*) from table1
where (
select
( (extract(day from column2)-extract(day from column1))*24*60 +
(extract(hour from column2)-extract(hour from column1))*60 +
extract(minute from column2)-extract(minute from column1)) as Processed_time from table1
) IN (1,5)
getting an error TOO MANY VALUES
View 2 Replies
View Related
Aug 17, 2012
/* Formatted on 08/17/2012 11:56:42 AM (QP5 v5.163.1008.3004) */
DECLARE
src_csr1 INTEGER;
src_x INTEGER;
dst_csr1 INTEGER;
v_DB_link VARCHAR2 (20);
v_Db_name VARCHAR2 (20);
v_sql1 VARCHAR2 (2000);
v_sql_src VARCHAR2 (2000);
[code]....
Getting the following error - I have checked and made sure that the query returns 4 columns.
DECLARE
*
ERROR at line 1:
ORA-00947: not enough values
ORA-06512: at line 60
View 8 Replies
View Related
