Heh... obviously, by the link you posted, you know some of the reasons why I avoid it.  Add to that that I have no need for the extended early dates...

• This reply was modified 7 years, 1 months ago by Jeff Moden.
• This reply was modified 1 years, 10 months ago by Jeff Moden. Reason: Cleanup code messed up due to forum "Upgrades"

Can't test it because I'm no where near a 2016 box just now but if all the dates in your JSON are in the DMY format, then just add the...

It's easy.  Just rename (re-alias) the column names and limit the rows with a WHERE clause.  Like this.

 SELECT  year1  = year3
        ,count1 = count3

By the same token, you can have a fire breathing monster of a server on a network with speeds that we could only dream of and still run into performance...

Nice article.  It would, however, be helpful if you would attach sample data in the "Resources" section so that people can "play" with your good code for additional learning through...

Just to establish some warm fuzzies that a concurrent update won't ever interfere with the correct calculation (unique sequential number) of the MatCodeID, I would add a WITH (TABLOCKX) to the...

Yes.  A "fact table" is a duplication of data that must be separately maintained.  It would probably be better to create what is known as a "Partitioned View" and reference that...

I see a claim of working against 1.7 million rows.  What does the data actually look like and what is the structure of the table that you're using for the...