percentage analysis in quartiles - oracle

I am trying to get the maximum, minimum of the difference between t3 and t3 in each quartile. But when I change the percentage # to 50,75.. I see the output listed below. I am not sure what I am doing wrong here.
SELECT MIN(SPREAD),MAX(SPREAD) FROM
(SELECT COUNT(*)*1.0 AS TOTALLINES FROM pa_fcs)A,
(SELECT SPREAD,ROWNUM *1.0 AS LINENUM FROM (
select BOARD_RATE_T3,BOARD_RATE_T1 ,(BOARD_RATE_T3-BOARD_RATE_T1) AS SPREAD
from pa_fcs ))B
WHERE LINENUM/TOTALLINES <= .25;
OUTPUT: .25
Min(Spread) Max(Spread)
7.47 5160.24
Output :.50
Min(Spread) Max(Spread)
7.47 5160.24
Output: .75
Min(Spread) Max(Spread)
0 5160.24


Use NTILE analytic function to calculate the quartiles.
Then is MIN, MAX trivial exersize
create table tab as
select rownum spread from dual connect by level <= 100;
with ntile as (
select
spread,
NTILE(4) OVER (ORDER BY spread DESC) qtile
from tab)
select qtile, min(spread), max(spread)
from ntile
group by qtile order by 1
;
QTILE MIN(SPREAD) MAX(SPREAD)
---------- ----------- -----------
1 76 100
2 51 75
3 26 50
4 1 25
For your table it would be something like
with spr as (
select (BOARD_RATE_T3-BOARD_RATE_T1) AS SPREAD
from pa_fcs),
ntile as (
select
spread,
NTILE(4) OVER (ORDER BY spread DESC) qtile
from spr)
select qtile, min(spread), max(spread)
from ntile
group by qtile order by 1

Related

Calculate current based on previous formula

I have a scenario where the current column value is calculated based on the previous value calculated by the formula
The initial row of the group has no previous value so it will not consider.
Formula for loss= relase-withdraw-least(previous_row_loss,reverse)
Here below loss is the column I need to calculate.
I tried with the following query but not getting expected output. Can you please guide me here.
SELECT
pid,release,withdraw,reverse,
SUM(release - withdraw - LEAST( LAG(loss,1,0) OVER (ORDER BY pid)),reverse)) as loss
FROM transactions

You can use a MODEL clause:
SELECT *
FROM (
SELECT t.*,
ROW_NUMBER() OVER (ORDER BY pid, fiscalperiod) AS rn
FROM table_name t
)
MODEL
DIMENSION BY (rn)
MEASURES (pid, fiscalperiod, release, withdraw, reverse, 0 AS loss)
RULES (
loss[1] = release[1] - withdraw[1] - reverse[1],
loss[rn>1] = release[cv()] - withdraw[cv()] - LEAST(reverse[cv()], loss[cv()-1])
+ loss[cv()-1]
);
Or, probably, much less efficiently a recursive query:
WITH numbered_rows AS (
SELECT t.*,
ROW_NUMBER() OVER (ORDER BY pid, fiscalperiod) AS rn
FROM table_name t
),
recursive_query (rn, pid, fiscalperiod, release, withdraw, reverse, loss) AS (
SELECT rn,
pid,
fiscalperiod,
release,
withdraw,
reverse,
release - withdraw - reverse
FROM numbered_rows
WHERE rn = 1
UNION ALL
SELECT n.rn,
n.pid,
n.fiscalperiod,
n.release,
n.withdraw,
n.reverse,
n.release - n.withdraw + GREATEST(r.loss - n.reverse, 0)
FROM numbered_rows n
INNER JOIN recursive_query r
ON (n.rn = r.rn + 1)
)
SELECT *
FROM recursive_query;
Which, for your sample data:
CREATE TABLE table_name (pid, fiscalperiod, release, withdraw, reverse) AS
SELECT 'A1', 2022001, 10, 10, 10 FROM DUAL UNION ALL
SELECT 'A1', 2022002, 20, 13, 2 FROM DUAL UNION ALL
SELECT 'A1', 2022003, 20, 20, 10 FROM DUAL UNION ALL
SELECT 'A2', 2022002, 15, 10, 13 FROM DUAL;
Both output:
RN
PID
FISCALPERIOD
RELEASE
WITHDRAW
REVERSE
LOSS
1
A1
2022001
10
10
10
-10
2
A1
2022002
20
13
2
7
3
A1
2022003
20
20
10
0
4
A2
2022002
15
10
13
5
db<>fiddle here

Oracle sort values into column

I have a table like this:
time length name
00:01:00 2 a
00:11:22 2 a
01:01:00 45 a
00:23:00 3 b
and I want to retrieve data from the table in the form:
a b
time length time length
00:01:00 2 00:23:00 3
00:11:22 2
01:01:00 2
so it is a simple task of rearranging data, atm I am doing this in a bash script, but I wonder if there is an easy way to do it in Oracle?

You can use analytical function ROW_NUMBER and full outer join as follows:
WITH CTE1 AS
(SELECT T.*, ROW_NUMBER() OVER (ORDER BY LENGTH, TIME) AS RN FROM YOUR_TABLE T WHERE NAME = 'a'),
CTE2 AS
(SELECT T.*, ROW_NUMBER() OVER (ORDER BY LENGTH, TIME) AS RN FROM YOUR_TABLE T WHERE NAME = 'b')
SELECT A.TIME, A.LENGTH, B.TIME, B.LENGTH
FROM CTE1 A FULL OUTER JOIN CTE2 B
ON A.RN = B.RN
Note: You need to use proper order by to order the records as per your requirement. I have used LENGTH, TIME

You can use a multi-column pivot, by adding an extra column that links the related A and B values; presumably by time order, something like:
select time_col, length_col, name_col,
dense_rank() over (partition by name_col order by time_col) as rnk
from your_table;
TIME_COL LENGTH_COL N RNK
-------- ---------- - ----------
00:01:00 2 a 1
00:11:22 2 a 2
01:01:00 45 a 3
00:23:00 3 b 1
and then pivot based on that:
select *
from (
select time_col, length_col, name_col,
dense_rank() over (partition by name_col order by time_col) as rnk
from your_table
)
pivot (
max(time_col) as time_col, max(length_col) as length_col
for name_col in ('a' as a, 'b' as b)
);
RNK A_TIME_C A_LENGTH_COL B_TIME_C B_LENGTH_COL
---------- -------- ------------ -------- ------------
1 00:01:00 2 00:23:00 3
2 00:11:22 2
3 01:01:00 45
I've left the rnk value in the output; if you don't want that you can list the columns in the select list:
select a_time_col, a_length_col, b_time_col, b_length_col
from ...
Or you could do the same thing with conditional aggregation (which is what pivot uses under the hood anyway):
select
max(case when name_col = 'a' then time_col end) as time_col_a,
max(case when name_col = 'a' then length_col end) as length_col_a,
max(case when name_col = 'b' then time_col end) as time_col_b,
max(case when name_col = 'b' then length_col end) as length_col_b
from (
select time_col, length_col, name_col,
dense_rank() over (partition by name_col order by time_col) as rnk
from your_table
)
group by rnk
order by rnk;
TIME_COL LENGTH_COL_A TIME_COL LENGTH_COL_B
-------- ------------ -------- ------------
00:01:00 2 00:23:00 3
00:11:22 2
01:01:00 45
db<>fiddle

3rd highest salary in oracle

I had been looking for the query to find the 3rd highest salary from the database (using Oracle database). I found the below query -
SELECT *
FROM
( SELECT e.*, row_number() over (order by sal DESC) rn FROM emp e
)
WHERE rn = 3;
I do not have oracle installed in my system, so I'm not try it out. But I want to know if the below query will work or not. If not, then why ?
WITH Sal_sort AS
(SELECT DISTINCT sal FROM salary ORDER BY sal DESC
)
SELECT * FROM Salary S, Sal_sort SS WHERE S.Sal = SS.Sal AND SS.rownum = 3;

Input Data
emp_no emp_fname emp_lname salary
1 aa bb 30
2 ee yy 31
3 rr uu 32
4 tt ii 33
5 tt ii 33
6 tt ii 33
7 tt ii 33
8 tt ii 30
9 tt ii 31
Example:
select * from ee;
select emp_no,salary ,dense_rank() over (order by salary ) dr
from ee
Output
emp_no salary dr
1 30 1
8 30 1
9 31 2
2 31 2
3 32 3
4 33 4
5 33 4
6 33 4
7 33 4

So much easier in version 12 of the database and higher now.
SELECT *
FROM employees
ORDER BY salary DESC OFFSET 2 ROWS FETCH NEXT 1 ROWS ONLY
Tim talks about this feature here.
And if you take a look at the plan, you can see it's not magic, the optimizer is using analytic functions to derive the results.

JUST ONE LINE
select * from (
select salary,dense_rank() over (order by salary desc) rank from employees) where rank=3;
or
select * from (
select a.*,dense_rank() over (order by a.salary desc) rank from employees a) where rank=3;

Without Dense_rank()
SELECT salary FROM employees
ORDER BY salary DESC
OFFSET 2
FETCH 1 NEXT ONE ROWS ONLY;
With Dense_rank()
SELECT salary
FROM
(
SELECT salary, DENSE_RANK() OVER (ORDER BY salary DESC) as rank from employees
)
WHERE rank = 3;

Issue in SQL query full scanning twice?

Table A
ID EmpNo Grade
--------------------
1 100 HIGH
2 105 LOW
3 100 MEDIUM
4 100 LOW
5 105 LOW
Query:
select *
from A
where EMPNO = 100
and rownum <= 2
order by ID desc
I tried this query to retrieve max and max-1 value; I need to compare the grade from max and max-1, if equals I need to set the flag as 'Y' or 'N' without using a cursor. Also I don't want to scan the entire record twice.
Please help me.

ROWNUM is applied before ORDER BY, so you need to nest the query like this:
select * from
(select * from A where EMPNO =100 order by ID desc)
where rownum<=2
That only performs one table scan (or it may use an index on EMPNO).

select *
from (
select id, emp_no, grade
, case
when lag(grade) over (order by emp_no desc) = grade
then 'Y'
else 'N'
end
as flag
, dense_rank() over( order by emp_no desc) as rank
from t
)
where rank <=2
;

Max size in a connected by prior Oracle

I've got some help turning my table of the sort:
Col
23
25
15
53
...
into something like 23,25,15,53...
The query that does it is
SELECT max(ltrim(sys_connect_by_path(flow_run_id, ','), ','))
FROM
(select flow_run_id, rownum rn
from table
where CREATED_DATE < sysdate - 32
and flow_id = 3
order by 1 desc)
START WITH rn = 1
CONNECT BY PRIOR rn = rn - 1
(this beaulty was given by Michael in here)
My current problem is that the result is too long (ORA-01489 over the 4k chars from varchar2). I'm still learning about these sys_connected_by_path so I'd need some help sorting this. How could I make this query return me multiple rows instead of one super long line? i.e.:
Instead of
419,1,2,3,411,418,4,415,887,413,414,201,888,890,401,417,610,412,416,5,6,922,1080,1422,1423,1411,1412,1413,1414,1415,1416,1417,1418,1419,1964,2217,1636,2037,1988,1970,2038,1989,2000,2040,1993,2043,1994,2001,2044,1658,1995,2045,2224,1996,2019,1678,1997,2022,2201,1680,2219,2024,2207,1677,2209,2220,1959,2211,1961,2026,2212,1962,2028,2215,1675,1676,2035,2216,1986,1963,2017,1983,1935,2002,2018,1985,1936,2003,2020,2032,1937,2004,2021,2033,1938,1943,2023,2034,1939,1944,2025,2225,1941,1950,2027,2036,1942,1955,2029,2041,1945,1956,2030,2227,1946,1957,2031,2039,1947,2005,1974,2042,1948,2006,1976,2228,1949,2007,1978,1951,2009,1979,1929,1952,2012,1980,1931,1953,2013,1981,1933,1954,2015,2334,2350,2311,2239,2240,2241,2242,2245,2246,2249,2250,2336,2312,2008,2010,2011,2014,2251,2253,2016,2243,2244,2247,2351,2248,(...)
get
419,1,2,3,411,418,4,415,887,413,414,201,888,890,401,417,610,412,416,5,6,922,1080
1423,1411,1412,1413,1414,1415,1416,1417,1418,1419,1964,2217,1636,2037,1988,1970,2038
2000,2040,1993,2043,1994,2001,2044,1658,1995,2045,2224,1996,2019,1678,1997,2022,2201
(...)
Any tips?
Thanks!
f.

the following query will cut your big string in parts:
SQL> SELECT root_rn, MAX(concat)
2 FROM (SELECT connect_by_root(rn) root_rn,
3 ltrim(sys_connect_by_path(flow_run_id, ','), ',') concat
4 FROM (SELECT flow_run_id, rownum rn
5 FROM (SELECT round(dbms_random.VALUE(1, 10000))
6 AS flow_run_id
7 FROM dual
8 CONNECT BY ROWNUM <= 2000)
9 ORDER BY 1 DESC)
10 START WITH MOD(rn, 10) = 1
11 CONNECT BY PRIOR rn = rn - 1
12 AND MOD(rn, 10) != 1)
13 GROUP BY root_rn
14 ORDER BY root_rn;
ROOT_RN MAX(CONCAT)
---------- -------------------------------------------------------------------
1 654,6710,5297,5481,5085,2793,7646,9170,1051,2387
11 1882,8285,5430,4928,267,3779,3843,1151,3085,1446
21 4721,6087,6755,9904,805,2776,4633,2772,7785,5818
31 5189,5307,6481,2099,3832,9788,5970,8068,6605,3904
41 53,7013,1314,7717,9320,7069,907,5367,5013,7637
51 3903,2318,2611,7954,5751,5598,6148,6555,9724,984
[...]
You can replace "10" with a bigger number if you want more elements on each row.

Some little modifications to keep order
SELECT 10*frn+1 root,ltrim(sys_connect_by_path(flow_run_id,','),',') FROM
(SELECT flow_run_id,mod(rn,10) mrn,floor(rn/10) frn,count(*)over(partition by floor(rn/10))-1 crn FROM
(SELECT flow_run_id, row_number()over(order by flow_run_id)-1 rn FROM
(SELECT round(dbms_random.VALUE(1, 10000)) AS flow_run_id FROM dual CONNECT BY ROWNUM <= 2000
)
)
)
WHERE crn = mrn
START WITH mrn = 0
CONNECT BY PRIOR mrn = mrn-1 AND PRIOR frn = frn

Resources