I have a table WHERE there's a column that has QTY Ranges like 1-5, 6-9 etc and prices in another column. i.e
Price QTY
------------------------
Price Qty Range
----- ----------
45 1-5
35 6-9
30 10-18
Now I want to get the result from the table where Qty is 7 Therefore Price returned should be 35 ( since Qty 7 falls in range 6-9
Any help greatly appriciated
Try this :-
Declare #val int
Set #val=7
;with cte(price,startVal,endVal) as
( Select price,
parsename(replace([Qty Range],'-','.'),2),
parsename(replace([Qty Range],'-','.'),1)
from yourTable
)
Select Price from cte
where #val between startVal and endVal
Result : 35
Demo in SQL FIDDLE
If you can't redesign the table to be sane, you can use a couple of CTEs to reconstruct it as a sane table for this query:
declare #PriceRanges table (Price int,QtyRange varchar(20))
insert into #PriceRanges (Price,QtyRange) values
(45,'1-5'),
(35,'6-9'),
(30,'10-18')
declare #Search int
set #Search = 7
;with FoundDashes as (
select Price,QtyRange,CHARINDEX('-',QtyRange) as DashPos
from #PriceRanges
)
, SaneRanges as (
select Price,CONVERT(int,SUBSTRING(QtyRange,1,DashPos-1)) as LowRange,CONVERT(int,SUBSTRING(QtyRange,DashPos+1,8000)) as HighRange
from FoundDashes
)
select Price from SaneRanges where #Search between LowRange and HighRange
Produces 35 as a result.
Related
I have a table of data which i am using a count statement to get the amount of records for the submission date
example
AuditId Date Crew Shift Cast ObservedBy 2ndObserver AuditType Product
16 2017-06-27 3 Day B1974, B1975 Glen Mason NULL Identification Billet
20 2017-06-29 1 Day 9879 Corey Lundy NULL Identification Billet
21 2017-06-29 4 Day T9627, T9625 Joshua Dwyer NULL ShippingPad Tee
22 2017-06-29 4 Day NULL Joshua Dwyer NULL Identification Billet
23 2017-06-29 4 Day S9874 Joshua Dwyer NULL ShippingPad Slab
24 2017-06-29 4 Day Bay 40 Joshua Dwyer NULL Identification Billet
Basically I am using the following code to get my results
SELECT YEAR([Date]) as YEAR, CAST([Date] as nvarchar(25)) AS [Date], COUNT(*) as "Audit Count"
FROM AuditResults
where AuditType = 'Identification' AND Product = 'Billet'
group by Date
this returns example
YEAR Date Audit Count
2017 2017-06-27 1
2017 2017-06-29 3
Now I want to be able to retrieve all dates even if blank
so I would like the return to be
YEAR Date Audit Count
2017 2017-06-27 1
2017 2017-06-28 0
2017 2017-06-29 3
I have the following function I am trying to use:
ALTER FUNCTION [dbo].[fnGetDatesInRange]
(
#FromDate datetime,
#ToDate datetime
)
RETURNS #DateList TABLE (Dt date)
AS
BEGIN
DECLARE #TotalDays int, #DaysCount int
SET #TotalDays = DATEDIFF(dd,#FromDate,#ToDate)
SET #DaysCount = 0
WHILE #TotalDays >= #DaysCount
BEGIN
INSERT INTO #DateList
SELECT (#ToDate - #DaysCount) AS DAT
SET #DaysCount = #DaysCount + 1
END
RETURN
END
How do I use my select statement with this function? or is there a better way?
cheers
Try this;
ALTER FUNCTION [dbo].[fnGetDatesInRange]
(
#FromDate datetime,
#ToDate datetime
)
RETURNS #YourData TABLE ([Year] int, DateText nvarchar(25),[Audit Count] int)
AS
begin
insert into #YourData
SELECT
YEAR(allDates.[Date]) as YEAR,
CAST(allDates.[Date] as nvarchar(25)) AS [Date],
COUNT(r.Product) as "Audit Count"
from
(
SELECT
[date]=convert(datetime, CONVERT(float,d.Seq))
FROM
(
select top 100000 row_number() over(partition by 1 order by A.name) as Seq
from syscolumns A, syscolumns B
)d
)allDates
left join
AuditResults r on r.[Date]=allDates.[date] and r.AuditType = 'Identification' AND r.Product = 'Billet'
where
allDates.[Date]>=#FromDate and allDates.[Date]<=#ToDate
group by
allDates.[Date]
return
end
The key is the 'allDates' section ;
SELECT
[date]=convert(datetime, CONVERT(float,d.Seq))
FROM
(
select top 100000 row_number() over(partition by 1 order by A.name) as Seq
from syscolumns A, syscolumns B
)d
This will return all dates between 1900 and 2173 (in this example). Limit that as you need but a nice option. A ton of different ways to approach this clearly
you have to create another table calendar as (Mysql)- idea is the same on all RDBMS-
CREATE TABLE `calendar` (
`dt` DATE NOT NULL,
UNIQUE INDEX `calendar_dt_unique` (`dt`)
)
COLLATE='utf8_general_ci'
ENGINE=InnoDB
;
and fill with date data.
more details
My data is like below:
ClassId ClassName StudentId Subject SubjectId
-----------------------------------------------------
1 ESL 12 English 20
1 ESL 13 Science 30
1 ESL 12 Social 40
1 ESL 12 Maths 50
Required output: parameters are Subject column values
ClassId ClassName TotalStudents SubjectIds
-----------------------------------------------
1 ESL 2 20, 40, 50, 30
When one student takes multiple subjects then count student only once, so in the above data 12 is one student id takes multiple subjects so counted only once. TotalStudents value is 2 (1 from student id 12 and 1 from student id 13)
I am not looking for how to display subjectIds column value in comma separated string.
Thanks in advance
COUNT DISTINCT then use STUFF for combined the subject
declare #temp table
(ClassId int,ClassName nvarchar(max),StudentId int,Subject nvarchar(max), SubjectId int)
insert into #temp values (1,'ESL',12,'English' , 20 )
insert into #temp values (1,'ESL',13,'Science' , 30 )
insert into #temp values (1,'ESL',12,'Social ' , 40 )
insert into #temp values (1,'ESL',12,'Maths ' , 50 )
select ClassId,ClassName,COUNT(DISTINCT StudentId) CNT,
STUFF( (SELECT ',' + CAST(t1.SubjectId AS NVARCHAR)
FROM #temp t1
WHERE StudentId = t1.StudentId
FOR XML PATH('')),
1, 1, '') SubjectIdS
from #temp
GROUP BY ClassId,ClassName
OUTPUT
DISTINCT can be applied inside aggregate functions.
SELECT COUNT(DISTINCT column_name) FROM table_name;
If you don't need to display the SubjectIds, then you need to use a GROUP BY clause to group the resultset by ClassId and ClassName.
SELECT ClassId, ClassName, COUNT(distinct StudentId) as TotalStudents
FROM MyTable
GROUP BY ClassId, ClassName
See this example at SqlFiddle
here is the dilemma i am having...
i have 2 tables
create table #orders
(orderNumber int,qty int,sku varchar(250),barcode varchar(250))
create table #allItemsInBox
([id] int,[date] date,[localsku] varchar(250),[box] varchar(250),barcode varchar(250))
i need to join the 2 tables on [barcode] and only have 1 result in the final table for every row in #allItemsInBox
please note [#allItemsInBox].[id] is unique the other fields in [#allItemsInBox] may not be
how would i go about doing something like this?
sample data:
[#orders]
(1,0,'10','10')
(1,0,'20','20')
(3,0,'20','20')
(4,0,'30','30')
(5,0,'40','40')
(6,0,'50','50')
#allItemsInBox
(1,'12/3/2014',10,'Box1',10)
(2,'12/2/2014',20,'Box2',20)
(3,'12/1/2014',20,'Box3',20)
(4,'11/30/2014',20,'Box4',20)
(5,'11/29/2014',30,'Box5',30)
(6,'11/28/2014',40,'Box6',40)
(7,'11/27/2014',60,'Box8',60)
(8,'11/27/2014',50,'Box10',50)
#output
(ordernumber int,uniqueitemID int,localsku varchar(250),box varchar(250))
(1,1,10,'Box1')
(1,2,20,'Box2')
(3,3,10,'Box3')
(4,5,30,'Box5')
(5,6,40,'Box6')
(6,8,50,'Box10')
This is quick but works. Depending on the size of your data this might be not the best way performance wise. But this will give you a start
DECLARE #orders TABLE (
orderNumber int,
qty int,
sku varchar(250),
barcode varchar(250)
)
DECLARE #allItemsInBox TABLE (
[id] int,
[date] date,
[localsku] varchar(250),
[box] varchar(250),
barcode varchar(250)
)
INSERT INTO #orders VALUES
(1,0,'10','10'),
(1,0,'20','20'),
(3,0,'20','20'),
(4,0,'30','30'),
(5,0,'40','40'),
(6,0,'50','50')
INSERT INTO #allItemsInBox VALUES
(1,'2014-12-03',10,'Box1',10),
(2,'2014-12-02',20,'Box2',20),
(3,'2014-12-01',20,'Box3',20),
(4,'2014-11-30',20,'Box4',20),
(5,'2014-11-29',30,'Box5',30),
(6,'2014-11-28',40,'Box6',40),
(7,'2014-11-27',60,'Box8',60),
(8,'2014-11-27',50,'Box10',50)
SELECT
orders.orderNumber AS ordernumber
,(SELECT TOP 1 allItems.id FROM #allItemsInBox allItems WHERE allItems.barcode = orders.barcode AND allItems.id >= orders.orderNumber ORDER BY allItems.id) AS uniqueitemID
,(SELECT TOP 1 allItems.localsku FROM #allItemsInBox allItems WHERE allItems.barcode = orders.barcode AND allItems.id >= orders.orderNumber ORDER BY allItems.id) AS localsku
,(SELECT TOP 1 allItems.box FROM #allItemsInBox allItems WHERE allItems.barcode = orders.barcode AND allItems.id >= orders.orderNumber ORDER BY allItems.id) AS box
FROM
#orders orders
Results in:
ordernumber uniqueitemID localsku box
1 1 10 Box1
1 2 20 Box2
3 3 20 Box3
4 5 30 Box5
5 6 40 Box6
6 8 50 Box10
edit: I updated the answer. You now have the same output as you specified in your example/question
I want to achieve something like;
eg.
ID VALUE NAME
1 1234 COUNTER
2 1235 BAGGAGE
OUTPUT:
COUNTER BAGGAGE
1234 1235
Script:
SELECT ID, VALUE, NAME FROM table WHERE ID = 1
UNION ALL
SELECT ID, VALUE, NAME FROM table WHERE ID = 2
RESULT:
ID VALUE NAME
1 1234 COUNTER
2 1235 BAGGAGE
Is this a row to column approach?
Another one standard solution
select max(case when name='counter' then value end) as counter,
max(case when name='baggage' then value end) as baggage
from table
In this case, you use the query below. It's not very generic but may be sufficient in specific cases as is simple and easy to use.
select
(select top 1 value from yourtable where name = 'counter') as counter,
(select top 1 value from yourtable where name = 'baggage') as baggage
SQL Fiddle demo
try this::
select id,case when name='counter' then value else ' ' end as counter,
case when name='baggage' then value else ' ' end as baggage
from table
it outputs
id counter baggage
1 1234
2 1235
fiddle demo
You could use a PIVOT like this:
DECLARE #DataSource TABLE
(
[ID] TINYINT
,[VALUE] SMALLINT
,[Name] VARCHAR(12)
)
INSERT INTO #DataSource ([ID],[VALUE],[Name])
VALUES (1,1234,'COUNTER')
,(2,1235,'BAGGAGE')
SELECT [COUNTER]
,[BAGGAGE]
FROM
(
SELECT [VALUE]
,[Name]
FROM #DataSource
) AS DS
PIVOT
(
MAX([VALUE]) FOR [Name] IN ([COUNTER], [BAGGAGE])
) AS PVT
OUTPUT:
Check the example.
Can someone steer me in the right direction for solving this issue with a set-based solution versus cursor-based?
Given a table with the following rows:
Date Value
2013-11-01 12
2013-11-12 15
2013-11-21 13
2013-12-01 0
I need a query that will give me a row for each date between 2013-11-1 and 2013-12-1, as follows:
2013-11-01 12
2013-11-02 12
2013-11-03 12
...
2013-11-12 15
2013-11-13 15
2013-11-14 15
...
2013-11-21 13
2013-11-21 13
...
2013-11-30 13
2013-11-31 13
Any advice and/or direction will be appreciated.
The first thing that came to my mind was to fill in the missing dates by looking at the day of the year. You can do this by joining to the spt_values table in the master DB and adding the number to the first day of the year.
DECLARE #Table AS TABLE(ADate Date, ANumber Int);
INSERT INTO #Table
VALUES
('2013-11-01',12),
('2013-11-12',15),
('2013-11-21',13),
('2013-12-01',0);
SELECT
DateAdd(D, v.number, MinDate) Date
FROM (SELECT number FROM master.dbo.spt_values WHERE name IS NULL) v
INNER JOIN (
SELECT
Min(ADate) MinDate
,DateDiff(D, Min(ADate), Max(ADate)) DaysInSpan
,Year(Min(ADate)) StartYear
FROM #Table
) dates ON v.number BETWEEN 0 AND DaysInSpan - 1
Next I would wrap that to make a derived table, and add a subquery to get the most recent number. Your end result may look something like:
DECLARE #Table AS TABLE(ADate Date, ANumber Int);
INSERT INTO #Table
VALUES
('2013-11-01',12),
('2013-11-12',15),
('2013-11-21',13),
('2013-12-01',0);
-- Uncomment the following line to see how it behaves when the date range spans a year end
--UPDATE #Table SET ADate = DateAdd(d, 45, ADate)
SELECT
AllDates.Date
,(SELECT TOP 1 ANumber FROM #Table t WHERE t.ADate <= AllDates.Date ORDER BY ADate DESC)
FROM (
SELECT
DateAdd(D, v.number, MinDate) Date
FROM
(SELECT number FROM master.dbo.spt_values WHERE name IS NULL) v
INNER JOIN (
SELECT
Min(ADate) MinDate
,DateDiff(D, Min(ADate), Max(ADate)) DaysInSpan
,Year(Min(ADate)) StartYear
FROM #Table
) dates ON v.number BETWEEN 0 AND DaysInSpan - 1
) AllDates
Another solution, not sure how it compares to the two already posted performance wise but it's a bit more concise:
Uses a numbers table:
Linky
Query:
DECLARE #SDATE DATETIME
DECLARE #EDATE DATETIME
DECLARE #DAYS INT
SET #SDATE = '2013-11-01'
SET #EDATE = '2013-11-29'
SET #DAYS = DATEDIFF(DAY,#SDATE, #EDATE)
SELECT Num, DATEADD(DAY,N.Num,#SDATE), SUB.[Value]
FROM Numbers N
LEFT JOIN MyTable M ON DATEADD(DAY,N.Num,#SDATE) = M.[Date]
CROSS APPLY (SELECT TOP 1 [Value]
FROM MyTable M2
WHERE [Date] <= DATEADD(DAY,N.Num,#SDATE)
ORDER BY [Date] DESC) SUB
WHERE N.Num <= #DAYS
--
SQL Fiddle
It's possible, but neither pretty nor very performant at scale:
In addition to your_table, you'll need to create a second table/view dates containing every date you'd ever like to appear in the output of this query. For your example it would need to contain at least 2013-11-01 through 2013-12-01.
SELECT m.date, y.value
FROM your_table y
INNER JOIN (
SELECT md.date, MAX(my.date) AS max_date
FROM dates md
INNER JOIN your_table my ON md.date >= my.date
GROUP BY md.date
) m
ON y.date = m.max_date