I have a table as Below....
ROW gvkey datadate CQTR CYEARQ Value
1 6066 3/31/2015 0:00 1 2015 3610
2 6066 12/31/2014 0:00 4 2014 16868
3 6066 9/30/2014 0:00 3 2014 10809
4 6066 6/30/2014 0:00 2 2014 6905
5 6066 3/31/2014 0:00 1 2014 3326
I want to get the sum of Value of 3/31/2015 and 12/31/2014. Please suggest how Can I do it in MS Sql.
Are you looking for this :-
Set Nocount On;
If Object_Id('tempdb.dbo.#table') Is Not Null
Begin
Drop Table #table;
End
Create Table #table
(
Id Int Primary Key
,Col1 Int
,RDate Datetime
,Col2 Int
,RYear Int
,Col3 Int
)
Insert Into #table Values
(1,6066,'03/31/2015 0:00',1,2015,3610)
,(2,6066,'12/31/2014 0:00',4,2014,16868)
,(3,6066,'09/30/2014 0:00',3,2014,10809)
,(4,6066,'06/30/2014 0:00',2,2014,6905)
,(5,6066,'03/31/2014 0:00',1,2014,3326)
Select t.Col1
,Sum(t.Col3) As ColSum
From #table As t With (Nolock)
Where t.RDate In ('03/31/2015','12/31/2014')
Group By t.Col1
Related
I have a table named three_current. this tables is inserted with 3 new rows every 1 minutes from another application, so the tables keeps on increasing in rows. These three new rows always have their channel number to be 350, 351, and 352. I want a trigger to insert each of these three rows into three separate tables such that each tables contains data with the same channel number.
The three_current tables is as such:
three_current table
datetime
channel_number
Value
Status
01/06/2021 22:45:00
350
100
1
01/06/2021 22:45:00
351
120
1
01/06/2021 22:45:00
352
110
1
01/06/2021 22:46:00
350
95
1
01/06/2021 22:46:00
351
105
1
01/06/2021 22:46:00
352
150
1
01/06/2021 22:47:00
350
195
1
01/06/2021 22:47:00
351
205
1
01/06/2021 22:47:00
352
250
1
I also have three other tables name red_current, yellow_current, and blue_current. I am trying without success to create a trigger to update these three tables based on the channel_number of three_current table such that
red_current table will be
datetime
channel_number
Value
Status
01/06/2021 22:45:00
350
100
1
01/06/2021 22:46:00
350
95
1
01/06/2021 22:47:00
350
195
1
yellow_current table will be
datetime
channel_number
Value
Status
01/06/2021 22:45:00
351
120
1
01/06/2021 22:46:00
351
105
1
01/06/2021 22:47:00
351
205
1
blue_current table will be
datetime
channel_number
Value
Status
01/06/2021 22:45:00
352
110
1
01/06/2021 22:46:00
352
150
1
01/06/2021 22:47:00
352
250
1
But what I get after executing my code is that the red_current, yellow_current and the blue_current tables are all being inserted with rows where the channel number is 350. This means that only the red_current table is correct while the other two tables are duplicates of the red_current table. (I feel my code can only execute for the first row of each updates received by three_current table and thats the row with channel number 350).
My code is as follows:
DELIMITER //
CREATE TRIGGER `add` AFTER INSERT ON `three_current`
FOR EACH ROW
BEGIN
DECLARE new_datetime datetime ; -- choose the datatypes
DECLARE new_channel_number int; --
DECLARE new_value double; --
DECLARE new_status smallint; --
SET new_datetime = new.datetime ;
SET new_channel_number = new.channel_number ;
SET new_value = new.value ;
SET new_status = new.status;
INSERT INTO red_current (datetime, channel_number, value, status)
SELECT new.datetime, new.channel_number , new.value, new.status
FROM three_current WHERE channel_number = '350'
ON DUPLICATE KEY UPDATE status = new.status;
INSERT INTO yellow_current (datetime, channel_number, value, status)
SELECT new.datetime, new.channel_number , new.value, new.status
FROM three_current WHERE channel_number = '351'
ON DUPLICATE KEY UPDATE status = new.status;
INSERT INTO blue_current (datetime, channel_number, value, status)
SELECT new.datetime, new.channel_number , new.value, new.status
FROM three_current WHERE channel_number = '352'
ON DUPLICATE KEY UPDATE status = new.status ;
END
//
DELIMITER ;
I am trying to make a calculation in MySQL for all columns of a table.
Table: bev
Jahr GKZ gesamt A B C
2017 1111000 88.519 855 888 814
2017 1112000 247.943 2.414 2.379 2.262
2017 1113000 253.106 2.290 2.343 2.289
2017 1113004 43.392 408 416 403
2017 1113008 12.383 137 134 124
2017 1113012 27.106 252 252 249
2017 1113016 41.673 391 410 398
2017 1113020 39.585 364 391 373
2017 1113024 10.075 63 73 74
2017 1113028 24.083 199 205 209
2017 1113032 8.745 63 77 65
2017 1113036 18.143 170 170 143
2017 1113040 27.921 243 215 251
Table: ja
GKZ Jahr ja_name
1001000 2017 K X
1002000 2017 K Y
5370000 2017 L Z
5370004 2017 Z1
5370012 2017 Z2
5370016 2017 Z3
5370020 2017 Z4
I already got the calculation for one column (the first one: gesamt) in a function:
CREATE DEFINER=`DB`#`%` FUNCTION `Total_Amount_Funct`(
bev_ID int(11),
bev_Total int(11),
ja_name VARCHAR(255),
ja_jahr int(11)) RETURNS int(11)
DETERMINISTIC
BEGIN
DECLARE Total_Amount int(11);
DECLARE kreis int(11);
DECLARE Total_Sum int(11);
SET kreis = (bev_ID / 1000) ;
SET Total_Sum = (SELECT SUM(b.gesamt)
FROM bev as b, ja as j
WHERE b.GKZ = j.GKZ
AND b.Jahr = j.Jahr
AND j.Jahr = ja_jahr
AND (MOD(b.GKZ, 1000) <> 0)
AND (MOD(b.GKZ, 1000) != 0)
AND NOT (MOD(b.GKZ, 1000) = 0)
AND (b.GKZ BETWEEN (kreis*1000 + 1) AND (((kreis+1)*1000)-1))
AND j.ja_name IS NOT NULL);
SET Total_Amount = bev_Total-Total_Sum;
RETURN (Total_Amount);
END
This function can be called with the following select:
SELECT DISTINCT
bev.GKZ,
bev.Jahr,
bev.gesamt,
CASE WHEN (bev.GKZ % 1000 = 0) THEN
coalesce(Total_Amount_Funct(bev.GKZ, bev.gesamt, ja.ja_name, bev.Jahr), bev.gesamt)
ELSE bev.gesamt
END AS bev,
ja.ja_name
FROM
ja, bev
WHERE
bev.GKZ = ja.GKZ
AND bev.Jahr = ja.Jahr;
I really would like to apply the function for all columns of the table. Maybe as a stored procedure? Maybe as dynamic columns. I do not know. I have solved this problem in MS SQL with dynamic columns but I have the feeling that translating it will take more time than trying to complete the function as a Stored Procedure.
The name of the columns can be obtained by:
SELECT column_name
FROM information_schema.columns
WHERE table_name='bev'
and column_name not in ('Jahr','GKZ');
As Result it should be:
GKZ Jahr gesamt bev ja_name
1111000 2017 88.519 88.519 K X
1112000 2017 247.943 247.943 K Y
1113000 2017 253.106 101.350 L Z
1113004 2017 43.392 43.392 Z1
1113012 2017 27.106 27.106 Z2
1113016 2017 41.673 41.673 Z3
1113020 2017 39.585 39.585 Z4
As you are using the column only in the SUM, you could pass the column name as parameter and use CASE-statement to pick the column accordingly. Something like:
CREATE FUNCTION `Total_Amount_Funct`(
bev_ID decimal(8,3),
bev_Total int,
ja_name VARCHAR(255),
ja_jahr int,
in_col varchar(10)
)
RETURNS int
DETERMINISTIC
BEGIN
DECLARE Total_Amount int(11);
DECLARE Total_Sum int(11);
SELECT
SUM(
case
when in_col='gesamt' then b.gesamt
when in_col='A' then b.A
when in_col='B' then b.B
when in_col='C' then b.C
end
) into Total_Sum
FROM bev as b
join ja as j on b.GKZ = j.GKZ AND b.Jahr = j.Jahr
WHERE
MOD(b.GKZ, 1000) != 0
AND b.GKZ BETWEEN bev_ID+1 AND bev_ID+999
AND j.ja_name IS NOT NULL
SET Total_Amount = bev_Total-Total_Sum;
RETURN (Total_Amount);
END
And then call the function with column name and correct value:
Total_Amount_Funct(bev.GKZ, bev.gesamt, ja.ja_name, bev.Jahr, 'gesamt'),
Total_Amount_Funct(bev.GKZ, bev.A, ja.ja_name, bev.Jahr, 'A')
...
Note that calling a function which makes a query will serialize your SQL (calling the function on each row causes the function query to be executed on each row). This will hurt the query performance.
slaakso,
thank you very much for your answer. You are from today my idol :-).
Thanks Thanks.
I have maybe one performance Question.
It is posible to write the function for all columns of the table bev. We can copy the column names in a temporary table:
CREATE TEMPORARY TABLE listColumns(
Columns_ID MEDIUMINT NOT NULL AUTO_INCREMENT ,
Columnsnamen varchar(256) ,
PRIMARY KEY (Columns_ID)
);
Readed from the System Information:
insert into listColumns (Columnsnamen)
SELECT column_name
FROM information_schema.columns
WHERE table_name='bev'
and column_name not in ('Jahr','GKZ');
This Table looks like:
Columns_ID Columnsnamen
1 gesamt
2 A
3 B
4 C
5
6
In such a way, that it is not necessary to mentione every column name (the table contains about 100 columns). Maybe with a cursor over the Columns_ID?
It woul be great if you have another advice for me.
Thank you and kind regads
Ana
Here's an example:
My database has several entries where the time values are:
10:00:00
10:10:00
10:20:00
...
Then sometimes there are these values where we have a 1 minute delay like:
10:20:00
10:31:00 <-
10:40:00
10:51:00 <-
And I wanted these numbers to, each, go to their respective minutes:
10:20:00
10:31:00 -> 10:30:00
10:40:00
10:51:00 -> 10:50:00
Edit: The time are sole entries, I have my date value in another column, for visualization my columns are
keyDate keyTime Sequency Value
12-03-2018 10:40:00 12 143
Please try this.
SELECT
CONVERT(TIME,CASE WHEN RIGHT(datepart(minute,BeginTime),1) = 1
THEN
CONVERT(VARCHAR(2),datepart(Hour,BeginTime)) +':'
+ CONVERT(VARCHAR(2),datepart(minute,BeginTime) - 1)
+':'
+ CONVERT(VARCHAR(2),datepart(second,BeginTime))
ELSE
CONVERT(VARCHAR(8),BeginTime)
END)
FROM #tbl
For example
Declare #tbl Table(
ArrangeId INT ,
BeginTime TIME,
EndDate NVARCHAR(50),
DeptId INT
)
INSERT INTO #tbl VALUES(1,'7:20:00','22:30',NULL)
INSERT INTO #tbl VALUES(3,'7:10:00','23:00',NULL)
INSERT INTO #tbl VALUES(2,'7:11:00','23:59',NULL)
INSERT INTO #tbl VALUES(4,'6:10:00','22:30',NULL)
INSERT INTO #tbl VALUES(4,'5:11:00','22:30',NULL)
INSERT INTO #tbl VALUES(4,'5:31:00','22:30',NULL)
SELECT
CONVERT(TIME,CASE WHEN RIGHT(datepart(minute,BeginTime),1) = 1
THEN
CONVERT(VARCHAR(2),datepart(Hour,BeginTime)) +':'
+ CONVERT(VARCHAR(2),datepart(minute,BeginTime) - 1)
+':'
+ CONVERT(VARCHAR(2),datepart(second,BeginTime))
ELSE
CONVERT(VARCHAR(8),BeginTime)
END)
FROM #tbl
Hi i need a complex query
my table structure is
attribute_id value entity_id
188 48,51,94 1
188 43,22 2
188 43,22 3
188 43,22 6
190 33,11 10
190 90,61 12
190 90,61 15
I need the count of the value like
attribute_id value count
188 48 2
188 43 3
188 51 1
188 94 1
188 22 2
190 33 1
190 11 1
190 90 2
190 61 2
I have searched a lot on google to have something like this but unfortunately i didn't get any success. Please suggest me how can i achieve this .
I use a UDF for things like this. If that could work for you:
CREATE FUNCTION [dbo].[UDF_StringDelimiter]
/*********************************************************
** Takes Parameter "LIST" and transforms it for use **
** to select individual values or ranges of values. **
** **
** EX: 'This,is,a,test' = 'This' 'Is' 'A' 'Test' **
*********************************************************/
(
#LIST VARCHAR(8000)
,#DELIMITER VARCHAR(255)
)
RETURNS #TABLE TABLE
(
[RowID] INT IDENTITY
,[Value] VARCHAR(255)
)
WITH SCHEMABINDING
AS
BEGIN
DECLARE
#LISTLENGTH AS SMALLINT
,#LISTCURSOR AS SMALLINT
,#VALUE AS VARCHAR(255)
;
SELECT
#LISTLENGTH = LEN(#LIST) - LEN(REPLACE(#LIST,#DELIMITER,'')) + 1
,#LISTCURSOR = 1
,#VALUE = ''
;
WHILE #LISTCURSOR <= #LISTLENGTH
BEGIN
INSERT INTO #TABLE (Value)
SELECT
CASE
WHEN #LISTCURSOR < #LISTLENGTH
THEN SUBSTRING(#LIST,1,PATINDEX('%' + #DELIMITER + '%',#LIST) - 1)
ELSE SUBSTRING(#LIST,1,LEN(#LIST))
END
;
SET #LIST = STUFF(#LIST,1,PATINDEX('%' + #DELIMITER + '%',#LIST),'')
;
SET #LISTCURSOR = #LISTCURSOR + 1
;
END
;
RETURN
;
END
;
The UDF takes two parameters: A string to be split, and the delimiter to split by. I've been using it for all sorts of different things over the years, because sometimes you need to split by a comma, sometimes by a space, sometimes by a whole string.
Once you have that UDF, you can just do this:
DECLARE #TABLE TABLE
(
Attribute_ID INT
,Value VARCHAR(55)
,Entity_ID INT
);
INSERT INTO #TABLE VALUES (188, '48,51,94', 1);
INSERT INTO #TABLE VALUES (188, '43,22', 2);
INSERT INTO #TABLE VALUES (188, '43,22', 3);
INSERT INTO #TABLE VALUES (188, '43,22', 6);
INSERT INTO #TABLE VALUES (190, '33,11', 10);
INSERT INTO #TABLE VALUES (190, '90,61', 12);
INSERT INTO #TABLE VALUES (190, '90,61', 15);
SELECT
T1.Attribute_ID
,T2.Value
,COUNT(T2.Value) AS Counter
FROM #TABLE T1
CROSS APPLY dbo.UDF_StringDelimiter(T1.Value,',') T2
GROUP BY T1.Attribute_ID,T2.Value
ORDER BY T1.Attribute_ID ASC, Counter DESC
;
I did an ORDER BY Attribute_ID ascending and then the Counter descending so that you get each Attribute_ID with the most common repeating values first. You could change that, of course.
Returns this:
Attribute_ID Value Counter
-----------------------------------
188 43 3
188 22 3
188 94 1
188 48 1
188 51 1
190 61 2
190 90 2
190 11 1
190 33 1
I want it to be in the format of "YYYY-###" where it gives the year and number ### incriminating starting from 1. ex: 2016-001, 2016-002, 2016-003,...
You can use a Numbers/Tally table if you like. I use a UDF to dynamically generate number ranges. As you can see in the function call I set the range from 1 to 12 with an increment of 1
Declare #Table table (SomeField int)
Insert into #Table values
(2015),(2016)
Select StringValue =cast(SomeField as varchar(25))+'-'+right('000'+cast(RetVal as varchar(25)),3)
,SomeField
,RetVal
From #Table A
Join (Select RetVal=cast(RetVal as int) from [dbo].[udf-Create-Range-Number](1,12,1)) B
on (1=1)
Returns
StringValue SomeField RetVal
2015-001 2015 1
2015-002 2015 2
2015-003 2015 3
2015-004 2015 4
2015-005 2015 5
2015-006 2015 6
2015-007 2015 7
2015-008 2015 8
2015-009 2015 9
2015-010 2015 10
2015-011 2015 11
2015-012 2015 12
2016-001 2016 1
2016-002 2016 2
2016-003 2016 3
2016-004 2016 4
2016-005 2016 5
2016-006 2016 6
2016-007 2016 7
2016-008 2016 8
2016-009 2016 9
2016-010 2016 10
2016-011 2016 11
2016-012 2016 12
The UDF
CREATE FUNCTION [dbo].[udf-Create-Range-Number] (#R1 money,#R2 money,#Incr money)
-- Syntax Select * from [dbo].[udf-Create-Range-Number](0,100,2)
Returns
#ReturnVal Table (RetVal money)
As
Begin
With NumbTable as (
Select NumbFrom = #R1
union all
Select nf.NumbFrom + #Incr
From NumbTable nf
Where nf.NumbFrom < #R2
)
Insert into #ReturnVal(RetVal)
Select NumbFrom from NumbTable Option (maxrecursion 0)
Return
End