.

One alternate approach could be to use OUTER APPLY and locate the logouts using SELECT TOP(1).  Then you could use ISNULL to assign a default value to cases where there's...

If you add GROUPING SETS the non dynamic code looks something like this.  Since most people want the total to appear on the bottom row, 2 grouping columns were added...

• This reply was modified 4 years, 7 months ago by Steve Collins.
• This reply was modified 4 years, 7 months ago by Steve Collins.

Ok nice.  Good.  This is my cup of tea.  I'll take a look and definitely reply.  Maybe a few days

cross apply (values (charindex(',', string),
                     len(string))) st(cloc,...

• This reply was modified 4 years, 7 months ago by Steve Collins.

Oops (again).  Thank you Jeff.  I updated the answer with before/after 2017.

It's too broad of a question question because everything depends on the totality of your situation.  In general imo clustering on a GUID (which is intentionally random) is likely to...

Rereading what both Jeffery and me posted, it's correct and doesn't need editing.  The CONVERT is not necessary.  Frederico is also correct too.  CONVERT will work if the correct format...

datetime DOES have a format on convert - and that is what is causing it to fail.

format 1 is US - so date string should be on format...

Datetime has no format so using CONVERT with a format code is not necessary.  The code should work if CONVERT is completely removed.  Or you could use CAST instead.

            
April 24, 2021 at 1:56 pm
            
#3875226

            
            
                                                
        

Not too long ago there was a question for SQL Server 2012 with pretty much the same pattern.

https://www.sqlservercentral.com/forums/topic/how-to-make-pivot-to-feature-values-based-on-part-ca-and-x-exist-on-input-data/#post-3822116

Yes, no STRING_AGG in 2012.  We're in the 2016 Forum 🙂  To revert you could use STUFF+XML instead of STRING_AGG.

Not dynamic

select tmp_channel, tmp_entity,
       sum(case when tmp_category='Cat1' then tmp_charges else 0 end) cat1,
       sum(case when...

• This reply was modified 4 years, 7 months ago by Steve Collins.

A man goes to the doctor and says, "Doc, it hurts when I do this."  The doctor's response: "Don't do that."  What about not deleting the data within the transaction?

I think you're looking for something like this.  The first CROSS APPLY 'cma_check(cnt)' counts the number of commas.  The second CROSS APPLY 'wrd_check(cnt)' counts the number of words in the...