Considering a table customerBalance with 3 columns: name, date and balance. Suppose a set of records like:
cus_name cus_date cus_balance
John 06/14/2011 1000
John 06/15/2011 500
John 06/16/2011 0
Mary 06/14/2011 3000
Mary 06/15/2011 2800
Mary 06/16/2011 0
How to create a SQL query which returns, for the date 6/16/2011 instead 0, the last non-zero value based on date (in sample, $500 for John and $2800 for Mary)?
I'm trying to do it using a subquery which uses Max function to retrieve the last date with non-zero value, but I didn't succeed. This example is quite "nonsensical", but I really need to do an operation like this in my dataset. Thanks!
Note: If you can specify the DB and version this query can be improved.
Try this:
SELECT *
FROM customers
WHERE (cus_name, cus_date)
IN
(
SELECT cus_name, MAX(cus_date)
FROM customers
WHERE cus_balance <> 0
GROUP BY cus_name
)
Update: Alternate version:
SELECT a.*
FROM customers a,
(
SELECT cus_name, MAX(cus_date)
FROM customers
WHERE cus_balance <> 0
GROUP BY cus_name
) b
WHERE a.cus_name = b.cus_name
AND a.cus_date = b.cus_date
Here it goes:
CREATE Table #temp
(
Cus_Name VARCHAR(200) NULL,
Cus_Date Char(8) NULL,
Cus_Balance INT NULL
)
INSERT INTO #temp VALUES ('John' , '20110614' ,1000 )
INSERT INTO #temp VALUES ('John' , '20110615' , 500 )
INSERT INTO #temp VALUES ('John' , '20110616' , 0 )
INSERT INTO #temp VALUES ('Mary' , '20110614' ,3000 )
INSERT INTO #temp VALUES ('Mary' , '20110615' ,2800 )
INSERT INTO #temp VALUES ('Mary' , '20110616' , 0 )
SELECT
T.Cus_Name ,
MIN(t.Cus_Balance)
FROM #temp t
WHERE t.Cus_Balance <>0
GROUP BY t.Cus_Name
DROP TABLE #temp
Related
I have a use case where there is a free text field and the user id in the format ab12345 (fixed) and name (dynamic) can appear anywhere in the string.
Now I need to replace the ab12345 with xxxxxxx and the names also with XXXX wherever I find them in the string.
I used:
select *
from dbo.TEST
WHERE DESCRIPTION like '%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%';
to get the user id ab12345 but I am unable to write the replace function for this since the result is dynamic.
Same with the name as well.
the following may help in redacting the userID
USE tempdb
GO
CREATE TABLE #CustComments
( CustomerID INT
, CustomerNotes VARCHAR(8000)
)
GO
INSERT dbo.#CustComments
( CustomerID
, CustomerNotes
)
VALUES
( 1, 'An infraction was raised on user id ab12345, and the name of the complainant is John')
, ( 2, 'The customer was not happy with person CD45678 and is going to ask William Jones to speak with George Hillman about this matter' )
, ( 3, 'A customer called and repeatedly mentioned the name of employee ZX98765 and assumes their name was Janet which is not correct')
SELECT * ,
PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes) start_pos,
SUBSTRING (customernotes, (PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes)) ,7 ) extractstring,
REPLACE(customernotes, substring (customernotes, (PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes)) ,7 ), 'XXXXXXX') redacted
FROM #CustComments
--TIDY UP
DROP TABLE #CustComments
if you have or can create a table of "names"...this may work
USE tempdb
GO
CREATE TABLE #CustComments (
CustomerID int,
CustomerNotes varchar(8000)
)
GO
INSERT #CustComments (CustomerID
, CustomerNotes)
VALUES (1, 'An infraction was raised on user id ab12345 , and the name of the complainant is Ann')
, (2, 'The customer was not happy with person CD45678 and is going to ask Richard Jones to speak with Todd Hillman about this matter')
, (3, 'A customer called and repeatedly mentioned the name of employee ZX98765 and assumes their name was Shana which is not correct')
CREATE TABLE #empname (
ename varchar(255) NOT NULL
)
GO
INSERT INTO #empname ([ename])
VALUES ('Zeph'), ('Ebony'), ('Felicia'), ('Benedict'), ('Ahmed'), ('Ira'), ('Julie'), ('Levi'),
('Sebastian'), ('Fiona'), ('Lamar'), ('Russell'), ('Abdul'), ('Lev'), ('Isaiah'), ('Charlotte'),
('Rowan'), ('Ivory'), ('Quinn'), ('Jordan'), ('Xantha'), ('Shana'), ('Mufutau'), ('Jessamine'),
('Desirae'), ('Yvette'), ('Odessa'), ('Ray'), ('Ori'), ('Zenaida'), ('Allegra'), ('Allistair'),
('Raymond'), ('Martena'), ('Cameron'), ('Ila'), ('Nigel'), ('Dale'), ('Emerald'), ('Guinevere'),
('Boris'), ('Dolan'), ('Ainsley'), ('Madeson'), ('Kadeem'), ('Ciaran'), ('Hop'), ('Louis'),
('Maia'), ('Hiroko'), ('Hakeem'), ('Cole'), ('Tyrone'), ('Amy'), ('Doris'), ('Keaton'),
('Carlos'), ('Richard'), ('Lysandra'), ('Beverly'), ('Hamish'), ('Demetria'), ('Eric'), ('Nayda'),
('Sydney'), ('Fritz'), ('Blaze'), ('Regina'), ('Ciara'), ('Ina'), ('Joan'), ('Risa'),
('Alea'), ('Denton'), ('Daryl'), ('Mollie'), ('Keane'), ('Jarrod'), ('Ann'), ('Juliet'),
('Germaine'), ('Alexa'), ('Zane'), ('Kiona'), ('Armand'), ('Jin'), ('Geraldine'), ('Natalie'),
('Nomlanga'), ('Todd'), ('Rajah'),('Lucian'), ('Idona'), ('Autumn'), ('Briar'),
-- add surname
('Hillman');
;
-- redact the userID std format
SELECT
CustomerID ,
--PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes) start_pos,
--SUBSTRING (customernotes, (PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes)) ,7 ) extractstring,
REPLACE(customernotes, substring (customernotes, (PATINDEX('%[a-zA-z][a-zA-Z][0-9][0-9][0-9][0-9][0-9]%', CustomerNotes)) ,7 ), 'XXXXXXX') ID_redacted
INTO #ID_REDACT
FROM #CustComments
-- split into rows
SELECT customerId, value
into #SPLIT
FROM #ID_REDACT
CROSS APPLY STRING_SPLIT(ID_redacted, ' ');
--redact based on join with a ""name"" table
SELECT s.customerid,
CASE
WHEN e.ename IS NULL THEN s.value
ELSE 'XXXXXXX'
END AS name_redact
INTO #NAME_REDACT
FROM #split AS s
LEFT OUTER JOIN #empname AS e
ON s.value = e.ename
SELECT customerId,
STRING_AGG(name_redact, ' ') as full_redact
INTO #RESULTS
from #NAME_REDACT
group by CustomerID
-- RESULTS WITH COMPARISON
SELECT
C.CustomerID,
C.CustomerNotes AS Original,
R.full_redact AS Redacted
FROM #CustComments AS C
INNER JOIN #RESULTS AS R
ON C.CustomerID = R.customerId
--TIDY UP
DROP TABLE #CustComments
DROP TABLE #empname
DROP TABLE #ID_REDACT
DROP TABLE #SPLIT
DROP TABLE #NAME_REDACT
DROP TABLE #RESULTS
I'm trying to create a report with one row per customer, however there are multiple rows per customer.
The current view is:
Customer Business Dept Type Status
-----------------------------------------------
019 Public null null null
019 null IT null null
019 null null Retail 0 --char(1)
My desired view is:
Customer Business Dept Type Status
-----------------------------------------------
019 Public IT Retail 0
I'm using SQL Server 2008 R2. There are more columns in my data set, but this is a sample. I'm unsure of how to achieve the results when my datatype is character and not INT based.
If this is a representative example, and each column will always have a single row with a value and the others will have nulls, you could use an aggregate max or min, which ignore nulls:
SELECT customer, MAX(business), MAX(dept), MAX(type), MAX(status)
FROM mytable
GROUP BY customer
try something this:
CREATE TABLE #tmp ([Customer] CHAR(3), [Business] VARCHAR(20), [Dept] VARCHAR(20), [Type] VARCHAR(20), [Status] CHAR(1))
INSERT INTO #tmp (Customer, Business) VALUES ( '019', 'Public')
INSERT INTO #tmp (Customer,Dept) VALUES ('019','IT')
INSERT INTO #tmp (Customer,[Type]) VALUES ('019','Retail')
INSERT INTO #tmp (Customer,[Status]) VALUES ('019','0')
SELECT * FROM #tmp AS t
SELECT t.Customer, t.Business, t2.Dept, t3.[Type], t4.[Status] FROM #tmp AS t
JOIN #tmp AS t2 ON t2.Customer = t.Customer
JOIN #tmp AS t3 ON t3.Customer = t.Customer
JOIN #tmp AS t4 ON t4.Customer = t.Customer
WHERE t.Business IS NOT NULL AND t2.Dept IS NOT NULL AND t3.[Type] IS NOT NULL AND t4.[Status] IS NOT NULL
I have a table like this
USER itemnumber datebought (YYYYmmDD)
a 1 20160101
b 2 20160202
c 3 20160903
d 4 20160101
Now I have to show the total number of items bought by each user after date 20160202 (2 february 2016)
I used
SELECT USER, COUNT(itemnumber)<br/>
FROM TABLE<br/>
WHERE datebought >= 20160202<br/>
GROUP BY USER<br>
It gives me results
b 1
c 1
but I want like this
a 0
b 1
c 1
d 0
Please tell me what is the most quick method / efficient method to do that ?
Try like this,
DECLARE #table TABLE
(
[USER] VARCHAR(1),
itemnumber INT,
datebought DATE
)
INSERT INTO #TABLE VALUES
('a',1,'20160101'),
('b',2,'20160202'),
('b',2,'20160202'),
('b',2,'20160202'),
('c',3,'20160903'),
('d',4,'20160101')
SELECT *
FROM #TABLE
SELECT [USER],
Sum(CASE
WHEN datebought >= '20160202' THEN 1
ELSE 0
END) AS ITEMCOUNT
FROM #TABLE
GROUP BY [USER]
Use this
SELECT USER, COUNT(itemnumber)
FROM TABLE
WHERE datebought >= 20160202
GROUP BY USER
Though this query won't be a good idea for the large amount of data:
SELECT USER, COUNT(itemnumber)
FROM TABLE
WHERE datebought >= 20160202
GROUP BY USER
UNION
SELECT DISTINCT USER, 0
FROM TABLE
WHERE datebought < 20160202
USE tempdb
GO
DROP TABLE test1
CREATE TABLE test1(a NVARCHAR(10), ino INT, datebought INT)
INSERT INTO dbo.test1
( a, ino, datebought )
VALUES ( 'a' , 1 , 20160101)
INSERT INTO dbo.test1
( a, ino, datebought )
VALUES ( 'b' , 2 , 20160202)
INSERT INTO dbo.test1
( a, ino, datebought )
VALUES ( 'c' , 3 , 20160903)
INSERT INTO dbo.test1
( a, ino, datebought )
VALUES ( 'd' , 4 , 20160101)
SELECT * FROM dbo.test1
SELECT a, COUNT(ino) OVER(PARTITION BY a) FROM dbo.test1
WHERE datebought>=20160202
UNION ALL
SELECT a, 0 FROM dbo.test1
WHERE datebought<20160202
ORDER BY a
Current
Name Quantity
---------------
Stella 2
Jennifer 2
Greg 3
Requested result
Name Quantity
---------------
Stella 1
Stella 1
Jennifer 1
Jennifer 1
Greg 1
Greg 1
Greg 1
How should I do it?
declare #T table
(
Name varchar(50),
Sales int
)
insert into #T values
('Stella', '2'),
('Jennifer', '2'),
('Greg', '3')
If the maximum value in the quantity column is known to be less than 32,767, you can use Recursion to generate numbers and join the Numbers to achieve your result.
/*******************************************
Max Recursion Count in SQL Server is 32767
Limitation of 32767 Numbers!
******************************************/
;WITH Numbers (Number) AS
(
SELECT 1
UNION ALL
SELECT 1 + Number FROM Numbers WHERE Number < 100
)
SELECT m.Name,
Quantity = 1
FROM MyTable m
JOIN #numbers n ON m.Quantity <= n.Number
OPTION (MAXRECURSION 32767);
Using recursion and borrowing Michael Fredrickson's setup code:
declare #T table (
Name varchar(50),
Sales int
)
insert into #T values ('Stella', '2')
insert into #T values ('Jennifer', '2')
insert into #T values ('Greg', '3')
-- Recursive verion
;with People (Name, Sales) as
(
select Name, Sales
from #T
union all
select Name, Sales - 1
from People
where Sales - 1 > 0
)
select Name, 1 as Quantity
from People
option (maxrecursion 0) -- Recurse without limit
This seems to run faster on my box (5x faster than Michael Fredrickson's according to query plan, but with many more logical reads), not that it matters much.
You'll probably want to have a pre-populated numbers table to do this:
declare #T table (
Name varchar(50),
Sales int
)
declare #numbers table (
Number int
)
insert into #numbers values (1)
insert into #numbers values (2)
insert into #numbers values (3)
insert into #numbers values (4)
-- Etc... up to however many numbers is the max possible value for sales...
insert into #T values ('Stella', '2')
insert into #T values ('Jennifer', '2')
insert into #T values ('Greg', '3')
SELECT
t.Name,
1 AS Sales
FROM
#T t JOIN
#numbers n ON
t.Sales >= n.Number
ORDER BY t.Name
That's how you could do it, but I'm not sure on why you would want to do it.
SELECT id, EmpNo
FROM EmployeesTable
EmpNo can be the same for 1 or more records in the results of the above query. I now want to add another column derived from EmpNo(lets call it EmpNo2) but only returning distinct values of EmpNo.
For example if the above query returns 100 records but there are 69 distinct EmpNo values and i modify the query to
SELECT id, EmpNo, Distinct EmpNo2
FROM EmployeesTable EmpNo
,
i want all the 100 rows to be returned but the last column EmpNo2 should return 69 distinct values of EmpNo field.
But as already know, using distinct in that way results into an error but i want to implement such functionality - and a subquery is not helping.
SAMPLE REQUIRED RESULTS
ID EmpNo EmpNo2
1 0T4/HR 0T4/HR
1 0T4/HR 2VP/E
1 0T4/HR xT9/67
1 0T4/HR
1 0T4/HR
2 2VP/E
2 2VP/E
2 2VP/E
2 2VP/E
2 2VP/E
3 XT9/67
3 XT9/67
3 xT9/67
3 XT9/67
How about:
Select id, empno, empno2
from employeestable left outer join (
SELECT min([id]) as minid
,[empno] empno2
FROM [EmployeesTable]
group by empno) etab2 on employeestable.id = etab2.minid
You're saying a subquery won't work, though - why not?
Your requirement is not clear and I also have very little information. Following is what you need. This can be even better but it is just a try.
declare #temp table
(
uniqueid int identity(1, 1),
id int,
empno varchar(50),
empno2 varchar(50)
)
insert into #temp select 1, '0T4/HR', null
insert into #temp select 1, '0T4/HR' , null
insert into #temp select 1 , '0T4/HR' , null
insert into #temp select 1, '0T4/HR' , null
insert into #temp select 1, '0T4/HR' , null
insert into #temp select 2, '2VP/E' , null
insert into #temp select 2, '2VP/E' , null
insert into #temp select 2, '2VP/E' , null
insert into #temp select 2, '2VP/E' , null
insert into #temp select 2, '2VP/E' , null
insert into #temp select 3, 'XT9/67' , null
insert into #temp select 3, 'XT9/67' , null
insert into #temp select 3, 'xT9/67' , null
insert into #temp select 3, 'XT9/67' , null
SELECT ROW_NUMBER() OVER (ORDER BY id) AS id, empno into #temp FROM #temp group by empno, id
update #temp set empno2 = t2.empno
from #temp t inner join #temp t2 on t.uniqueid = t2.id
select * from #temp
drop table #temp