Sleep

A way to get more accurate results is to use the Advanced tab from the flat file connection.

There's a button that says "Suggest Types...", there you can tell the number...

I repeat, change your AND into an OR.

Or change the time to make them all in the same day.

selectWeekEnding,

count(*) as TotalVisits_3To2

--into #WeekEndingCountTotal_3To2

from #Test t

cross apply (

...

How would I know? You're the one asking for "patients that were registered between 3PM and 2AM the next day."

I assumed you...

Something like this?

DECLARE @Date datetime = '20161002';

SELECT *

FROM #Test

WHERE RegistrationDateTime >= DATEADD( HH, 15, @Date)

AND RegistrationDateTime < DATEADD( HH, 26, @Date)

ORDER BY RegistrationDateTime

Exactly what Tim said. Just be aware that the tables' structure should be the same on all connections or you'll be prone to errors.

Did you create the import from scratch, again? It might have kept the length in the package even if the column length was increased.

You might just need to change "digit / 10" to "1".

This might work.

create table #mytab (origval varchar(max))

insert #mytab values

('1.2'), --should convert to 1.3

('6.25'), --should convert to...

Minnie

Here's a different option:

selecT uc.UserGroup, us.Item

FROM dbo.DelimitedSplit8K(@UserGroup,'^') gs

CROSS APPLY (SELECT LEFT( gs.Item, NULLIF(CHARINDEX( '@', gs.Item), 0) - 1) AS UserGroup,

...

Maybe something like this could help:

DECLARE @sql nvarchar(MAX)

SET @sql = ( SELECT N'EXECUTE sp_order_historay ' + CAST( orderID AS nvarchar(10)) + N';' + NCHAR(10)

...

Don't think of tables. You need to start by thinking on entities, attributes and relationships. Create an ER model before even thinking about tables.

Designing a DB is not a task...

Zeus

Here's an option:

Declare @UserGroup nvarchar(MAX);

SET @UserGroup = 'UserGroup1|User1|User2|User3&UserGroup2|user4|user5|user6&Usergroup3|user7|user8|user9';

selecT uc.UserGroup, us.Item

FROM dbo.DelimitedSplit8K(@UserGroup,'&') gs

CROSS APPLY (SELECT LEFT( gs.Item, CHARINDEX( '|', gs.Item + '|') - 1) AS UserGroup,

...