A good place to start could be with a query to generate the most recent 5 Friday dates...

Here are two ways.  My preference is for the second one.  Since the FORMAT function was introduced in SQL Server 2012 (afaik) it's been known to have performance issues.  Maybe...

IMHO, Steve Collins' post (with a tiny tweak (change fn.n to fn.value) in the 2nd bit of code) contains the winning entries for simplicity IF you don't mind any...

Using a tally function or GENERATE_SERIES maybe something like this

/* expand the date range using a tally function */
select v.calc_dt, count(*) active_policies
from #Policy p
   ...

A while ago when I looked into GitHub Copilot, the service was based on ChatGPT-3.5 which is/was quite inferior imo to the ChatGPT-4 which was available at the time. 

Back in the day, before SQL Server 2012 was previewed, I worked on a project which started with a blank database.  My teammate at the time started sending me articles...

We'll have to see how the OP replies.  Maybe my code only summed the part they want subtracted from the sum_qty

All right, some data.  This could be my only chance to use SQL for the day.  For my part, this week and until it's done, I'm working on web development. ...

That old function could use an updating imo.  As this is the 2022 topic GENERATE_SERIES and STRING_AGG are available.  Also, not to say it's incorrect but...

To make the selection criteria specific to each column you could try a conditional aggregation approach to pivoting the data

select TestID,
      ...

This question still has me thinking.  How has development changed because of AI?  There's more than one trend happening at once imo.  For SQL the impact for me has been...

Thanks alot guys, I think I've taken a bit from each of the 3 posts and have made progress. I've posted the new execution plan below. Seems to be...

Disclaimer: this contains speculation and opinions

In broad terms, when considering performance, imo it makes sense to overlay the FROM clause with the execution plan.  The FROM clause is read left-to-right...

This query looks for the minimum previous appropriate enter_date for each row and then calculates the DATEDIFF.  The sample was expanded to include a greater than 12 month gap

Aus dem Paradies, das Cantor uns geschaffen, soll uns niemand vertreiben können