How to select distinct multi-column values in Oracle SQL? - sql

I am trying to get distinct values with multi column select.
Sample table:
CREATE TABLE DUP_VALUES (ID NUMBER, NAME VARCHAR2(64));
INSERT INTO DUP_VALUES values (1, 'TEST1');
INSERT INTO DUP_VALUES values (2, 'TEST1');
INSERT INTO DUP_VALUES values (3, 'TEST2');
INSERT INTO DUP_VALUES values (4, 'TEST2');
INSERT INTO DUP_VALUES values (5, 'TEST1');
INSERT INTO DUP_VALUES values (6, 'TEST1');
INSERT INTO DUP_VALUES values (7, 'TEST1');
I want to get
ID NAME
1 TEST1
3 TEST2
I tried with SELECT DISTINCT ID, NAME FROM DUP_VALUES
But, I got all values, because ID is unique.


Use aggregation:
select min(id) as id, name
from dup_values
group by name;

Related

How to calculate the average value of following n rows based on another column - SQL (Oracle)

I am trying to calculate average monthly value of premiums for each POLICY_ID in monthly basis as shown below. When a customer updates his/her yearly payment frequency to a value different than 12, I need to manually calculate the average monthly value for the PREMIUM. How can I achieve the values shown in MONTHLY _PREMIUM_DESIRED?
Thanks in advance.
Note: Oracle version 12c
What I've tried:
SELECT
T.*,
SUM(PREMIUM) OVER(PARTITION BY T.POLICY_ID ORDER BY T.POLICY_ID, T.PAYMENT_DATE ROWS BETWEEN CURRENT ROW AND 12/T.YEARLY_PAYMENT_FREQ-1 FOLLOWING ) / (12/T.YEARLY_PAYMENT_FREQ) MONTLY_PREMIUM_CALCULATED
FROM MYTABLE T
;
Code for data:
DROP TABLE MYTABLE;
CREATE TABLE MYTABLE (POLICY_ID NUMBER(11), PAYMENT_DATE DATE, PREMIUM NUMBER(5), YEARLY_PAYMENT_FREQ NUMBER(2),MONTHLY_PREMIUM_DESIRED NUMBER(5));
INSERT INTO MYTABLE VALUES (1, DATE '2014-10-01',120,12,120);
INSERT INTO MYTABLE VALUES (1, DATE '2014-11-01',360,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2014-12-01',0,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-01-01',0,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-02-01',360,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-03-01',0,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-04-01',0,4,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-05-01',720,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-06-01',0,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-07-01',0,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-08-01',0,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-09-01',0,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-10-01',0,2,120);
INSERT INTO MYTABLE VALUES (1, DATE '2015-11-01',120,12,120);
INSERT INTO MYTABLE VALUES (2, DATE '2015-01-01',60,3,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-02-01',0,3,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-03-01',0,3,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-04-01',0,3,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-05-01',180,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-06-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-07-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-08-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-09-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-10-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-11-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2015-12-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-01-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-02-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-03-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-04-01',0,1,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-05-01',15,12,15);
INSERT INTO MYTABLE VALUES (2, DATE '2016-06-01',15,12,15);
SELECT * FROM MYTABLE;
EDIT:
Regardless from payment frequency PREMIUM amount can also be changed by customer. Below, for the POLICY_ID = 1, I have added new records starting from "2015/11/01" to demonstrate this situation. In this case, average monthly premium increased from 120 to 240.
Also removed the screenshot to make the question more readable.
DROP TABLE MYTABLE2;
CREATE TABLE MYTABLE2 (POLICY_ID NUMBER(11), PAYMENT_DATE DATE, PREMIUM NUMBER(5), YEARLY_PAYMENT_FREQ NUMBER(2),MONTHLY_PREMIUM_DESIRED NUMBER(5));
INSERT INTO MYTABLE2 VALUES (1, DATE '2014-10-01',120,12,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2014-11-01',360,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2014-12-01',0,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-01-01',0,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-02-01',360,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-03-01',0,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-04-01',0,4,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-05-01',720,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-06-01',0,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-07-01',0,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-08-01',0,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-09-01',0,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-10-01',0,2,120);
INSERT INTO MYTABLE2 VALUES (1, DATE '2015-11-01',240,12,240);
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-12-01',240,12,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-01-01',960,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-02-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-03-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-04-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-05-01',960,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-06-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-07-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (1, DATE '2016-08-01',0,4,240); --newly added records
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-01-01',60,3,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-02-01',0,3,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-03-01',0,3,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-04-01',0,3,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-05-01',180,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-06-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-07-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-08-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-09-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-10-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-11-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2015-12-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-01-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-02-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-03-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-04-01',0,1,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-05-01',15,12,15);
INSERT INTO MYTABLE2 VALUES (2, DATE '2016-06-01',15,12,15);
SELECT * FROM MYTABLE2;

I think the calculation is:
select t.*,
premium / (12 / yearly_payment_freq)) as monthly_premium_calculated
from mytable t;
EDIT:
I see, you also need this spread over the intermediate months. So you can assign the groups by counting the number of non-zero payments. Then:
select t.*,
( max(premium) over (partition by policy_id, grp) /
(12 / yearly_payment_freq)
) as monthly_premium_calculated
from (select t.*,
sum(case when premium > 0 then 1 else 0 end) over (partition by policy_id order by payment_date) as grp
from mytable t
) t;
Here is a db<>fiddle (it uses Postgres because that is easier to set up than Oracle).

SQL Server: Find the group of records existing in another group of records

I'm new to SQL Server and I searched for a solution to find, if a group is included in another group.
The query result should be grp_id 2 because 'A'+'B' is included in grp 3 and 5.
The result should be the grp_id of the the groups, that are included in other groups. With this result i´ll make an update of another table, joined with the grp_id.
The result should be:
+----+
| id |
+----+
| 2 |
+----+
I stuck in SQL because I do not find a solution to compare the groups. The idea was using bitwise comparison. But for that I had to add the value of each item in a field. I think there could be an easier way.
Thank you and best regards!
Eric
create table tmp_grpid (grp_id int);
create table tmp_grp (grp_id int, item_val nvarchar(10));
insert into tmp_grpid(grp_id) values (1);
insert into tmp_grpid(grp_id) values (2);
insert into tmp_grpid(grp_id) values (3);
insert into tmp_grpid(grp_id) values (4);
insert into tmp_grpid(grp_id) values (5);
--
insert into tmp_grp(grp_id, item_val) values (1, 'A');
insert into tmp_grp(grp_id, item_val) values (2, 'A');
insert into tmp_grp(grp_id, item_val) values (2, 'B');
insert into tmp_grp(grp_id, item_val) values (3, 'A');
insert into tmp_grp(grp_id, item_val) values (3, 'B');
insert into tmp_grp(grp_id, item_val) values (3, 'C');
insert into tmp_grp(grp_id, item_val) values (4, 'A');
insert into tmp_grp(grp_id, item_val) values (4, 'C');
insert into tmp_grp(grp_id, item_val) values (4, 'D');
insert into tmp_grp(grp_id, item_val) values (5, 'A');
insert into tmp_grp(grp_id, item_val) values (5, 'B');
insert into tmp_grp(grp_id, item_val) values (5, 'E');

Geez!
Technically speaking, group one is found in all other groups right? So, first a cross join to itself would be best with the condition that the values are the same AND that the groups are different, but before we do that we need to know how many items belong to each group so that's why we have the first select as a group that includes the count of elements per group, then join that with the cross join...Hope this helps.
select distinct dist_grpid
from
(select grp_id, count(*) cc from tmp_grp group by grp_id) g
inner join
(
select dist.grp_id dist_grpid, tmp_grp.grp_id, count(*) cc
from
tmp_grp dist
cross join tmp_grp
where
dist.item_val = tmp_grp.item_val and
dist.grp_id != tmp_grp.grp_id
group by
dist.grp_id,
tmp_grp.grp_id
) cj on g.grp_id = cj.dist_grpid and g.cc = cj.cc

Sample observations per group without replacement in SQL

Using the provided table I would like to sample let's say 2 users per day so that users assigned to the two days are different. Of course the problem I have is more sophisticated, but this simple example gives the idea.
drop table if exists test;
create table test (
user_id int,
day_of_week int);
insert into test values (1, 1);
insert into test values (1, 2);
insert into test values (2, 1);
insert into test values (2, 2);
insert into test values (3, 1);
insert into test values (3, 2);
insert into test values (4, 1);
insert into test values (4, 2);
insert into test values (5, 1);
insert into test values (5, 2);
insert into test values (6, 1);
insert into test values (6, 2);
The expected results would look like this:
create table results (
user_id int,
day_of_week int);
insert into results values (1, 1);
insert into results values (2, 1);
insert into results values (3, 2);
insert into results values (6, 2);

You can use window functions. Here is an example . . . although the details do depend on your database (functions for random numbers vary by database):
select t.*
from (select t.*, row_number() over (partition by day_of_week order by random()) as seqnum
from test t
) t
where seqnum <= 2;

How to insert with loops in postgresql

I am trying to insert a number of rows to a table where values came from a select query
insert into article_rel (x_id, y_id) VALUES (12, (SELECT id from table where name = 'string')
this should be the same as executing this
insert into article_rel (x_id, y_id) VALUES (12, 1)
insert into article_rel (x_id, y_id) VALUES (12, 3)
insert into article_rel (x_id, y_id) VALUES (12, 4)
and so on.
I saw this How to use a SQL for loop to insert rows into database? but I'm not sure how can this help me
Thanks in advance,

Use this query:
INSERT INTO article_rel (x_id, y_id)
SELECT 12, id
FROM table_name
WHERE name = 'string'

Sql Server query with date filter

I have a table like this;
ID int,
OrderedDate DateTime
I want to select only records of followed month.
For example result set:
ID OrderedDate
110 January
110 February
200 January
200 February
How can I write this query?

I think you want list of months that ID has orders in but with the months sorted by the month number instead of the name?
create table test21210
(
id int,
OrderedDate datetime
)
go
insert test21210 (id, OrderedDate) values (110, '1/1/2010')
insert test21210 (id, OrderedDate) values (110, '1/5/2010')
insert test21210 (id, OrderedDate) values (110, '1/10/2010')
insert test21210 (id, OrderedDate) values (110, '2/2/2010')
insert test21210 (id, OrderedDate) values (110, '2/4/2010')
insert test21210 (id, OrderedDate) values (110, '2/6/2010')
insert test21210 (id, OrderedDate) values (200, '1/3/2010')
insert test21210 (id, OrderedDate) values (200, '1/5/2010')
insert test21210 (id, OrderedDate) values (200, '1/7/2010')
insert test21210 (id, OrderedDate) values (200, '1/9/2010')
insert test21210 (id, OrderedDate) values (200, '2/3/2010')
insert test21210 (id, OrderedDate) values (200, '2/5/2010')
insert test21210 (id, OrderedDate) values (200, '2/7/2010')
insert test21210 (id, OrderedDate) values (200, '2/9/2010')
go
with idmonth (id, MonthNumber) as
(
select id, MONTH(ordereddate) as 'MonthNumber'
from test21210
group by id, MONTH(ordereddate)
)
select id, DATENAME(MONTH, STR(MonthNumber)+'/1/2000')
from idmonth
order by id, MonthNumber

The question seems a bit unclear. But the example makes it look like you are wanting to sort by ID then by month name. If so, then I think this will do it. I don't have SQL Server to test it, so I'm sure it has syntax or other errors.
SELECT ID, DATENAME(month, OrderedDate) AS OrderedDate from table
ORDER BY 1, MONTH( OrderedDate )