UPDATE Table1
SET [Marks] =
(
SELECT
CASE STATEMENTS
FROM Table2 T2
WHERE Table1.ID = T2.ID)
)
The above UPDATE statements works fine, but if the ID doesn't match then it insert NULL value for 'Marks'.
But i wanted to keep the original value for Marks in Table1 if the Table1 and Table2 ID doesn't match.
How do i implement that in my code please.
i also tried using WHERE EXISTS BUT STILL no luck. I wonder whats the exact use of it.
Any help much appreciated.
UPDATE Table1
SET [Marks] =
(
SELECT
CASE STATEMENTS
FROM Table2 T2
WHERE Table1.ID = T2.ID)
)
WHERE id IN (SELECT id FROM table2)
Related
I need to update a second table with the results of this query:
SELECT Tag, battery, Wearlevel, SensorTime
FROM (
SELECT m.* , ROW_NUMBER() OVER (PARTITION BY TAG ORDER BY SensorTime DESC) AS rn
FROM [dbo].[TELE] m
) m2
where m2.rn = 1;
But. I had a hard time fixing the SET without messing it up. I want to have a table which has all data from last date of each TAG without duplicates.
Below code maybe you want.
UPDATE
Table_A
SET
Table_A.Primarykey = 'ss'+Table_B.Primarykey,
Table_A.AddTime = 'jason_'+Table_B.AddTime
FROM
Test AS Table_A
INNER JOIN UsersInfo AS Table_B
ON Table_A.id = Table_B.id
WHERE
Table_A.Primarykey = '559713e6-0d85-4fe7-87a4-e9ceb22abdcf'
For more details, you also can refer below posts and blogs.
1. How do I UPDATE from a SELECT in SQL Server?
2. How to UPDATE from SELECT in SQL Server
I got this statement, which works in Oracle:
update table a set
a.attribute =
(select
round(sum(r.attribute1),4)
from table2 p, table3 r
where 1 = 1
and some joins
)
where 1 = 1
and a.attribute3 > 10
;
Now I would like to do the same statement in Exasol DB. But I got error [Code: 0, SQL State: 0A000] Feature not supported: this kind of correlated subselect (Session: 1665921074538906818)
After some research, I found out you need to write the query in following syntax:
UPDATE table a
set a.attribute = r.attribute2
FROM table a, table2 p, table3 r
where 1 = 1
and some joins
and a.attribute3 > 10;
The problem is I can't take sum of r.attribute2. So I get unstable set of rows. Is there any way to do the first query in Exasol DB?
Thanks for help guys!
Following SQL UPDATE statement will work for cases if JOIN between table1 and table2 are 1-to-1 (or if there is a 1-to-1 relation between target table and resultset of JOINs)
In this case target table val column is updated otherwise an error is returned
UPDATE table1 AS a
SET a.val = table2.val
FROM table1, table2
WHERE table1.id = table2.id;
On the other hand, if the join is causing multiple returns for single table1 rows, then the unstable error raised.
If you want to sum the column values of the multiplying rows, maybe following approach can help
First sum all rows of table2 in bases of table1 and use this sub-select as a new temp table, then use this in UPDATE FROM statement
UPDATE table1 AS a
SET a.val = table2.val
FROM table1
INNER JOIN (
select id, sum(val) val from table2 group by id
) table2
ON table1.id = table2.id;
I tried to solve the issue using two tables
In your case probably you will use table2 and table3 in the subselect statement
I hope this is the answer you were looking for
I am trying to find the sum of the values corresponding to a key that is present in Table 1 but not in Table 2.
These tables have been created based on some filters and represent the values on 2 different dates.
The two different dates are chosen from 2 different date tables which have an inactive relationship, specially created for this purpose.
I want to create a measure that finds out the sum.
Below is the syntax I have used:
Difference =
VAR
Table1 = CALCULATETABLE(VALUES('TableA'[Id]), 'TableA'[Type] = "ABC", ALL('Date'), USERELATIONSHIP('Date'[As of Date], Previous_Date[Previous_Date]),USERELATIONSHIP('Date'[As of Date], 'TableA'[As of Date]))
VAR
Table2 = CALCULATETABLE(VALUES('TableA'[Id]), 'TableA'[Type] = "ABC", USERELATIONSHIP('Date'[As of Date], 'TableA'[As of Date]))
RETURN
IF(AND(VALUES('TableA'[Id]) IN Table1 , NOT(VALUES('TableA'[Id])) IN Table2),
CALCULATE(SUM('TableA'[Values])),0)
It is error-free. However, when I drop the measure on a KPI visual, I am getting the following message:
Please tell me what is wrong with the syntax. Also, please let me know if there is any better code that can be written.
Kindly help.
The reason you are getting an error is that it expects a single value for x when you write x IN Table1 but using the VALUES function can return a list rather than a single value.
I'd try something more like this after the RETURN:
SUMX (
'TableA',
IF ( 'TableA'[Id] IN Table1 && NOT ( 'TableA'[Id] IN Table2 ), 'TableA'[Value] )
)
This iterates through each row of TableA and checks if the Id value in each row is in the tables you calculated. If the condition is met, it adds the Value. Otherwise IF returns a blank (which is treated the same as 0 in a sum) since there is no third argument.
I have two tables like below.
Profile : ID
Charac : ID, NAME, DATE
With the above tables, I am trying to get NAME from Charac where we have max date.
I am trying to do a join with proc sql by replicating the answer for mysql like below
proc sql;
create table ggg as
select profile.ID ,T2.NAME
from Profile
left join
( select ID,max(DATE) as max_DATE
from EDW.CHARAC
group by ID
) as T1
on fff.ID = EDW.ID
left join EDW.CHARAC as T2
on T2.ID = T1.max_DATE
order by profile.ID DESC;
quit;
Error
ERROR: Unresolved reference to table/correlation name EDW.
ERROR: Expression using equals (=) has components that are of different data types.
Could it be you intended
on T2.ID = T1.max_DATE
which is probably source of "components that are of different data types" error
to be:
on T2.ID = T1.ID and T2.DATE = T1.max_DATE
that, is - joining on IDs at maximum DATE?
You can't use EDW like that. You need to join
on fff.ID=T1.ID
As far as data types, that probably is because EDW.ID is undefined and thus numeric by default.
I'm doing a count from table1 whose records/rows don't exist in table2
Here is the query:
select count(1) from table1
where not exists (select 1 from table2 where
table1.col1 = table2.col1
and table2.id=1)
I need to see the records that are missing in table2 , whose id in table2=1, and these records should be available in table1. The PK here is col1.
The query returns me 0. But if I do an excel sheet comparing by removing both the tables to excel. I can find 1591 records that are missing from table1 and are available in table2.
Your query is working fine.
You query finds records that EXISTS in table1 but not in table2
You have found with excel records that does NOT EXISTS in table1 and EXISTS in table2
If you'd like to find these records with SQL than your query should be:
select count(1) from table2
where table2.id=1 and table2.col1 not in (select col1 from table1)
or with not exists version of this query:
select count(1) from table2
where table2.id=1 and
not exists (select 1 from table1 where table1.col1=table2.col1)
I didn't test the queries.