Hi, mostly just curious here as both work, but also wanting to optimize for performance:
I have a Projects table with a row for every project status mutation. So a project can be New, go into Quotation, go into Accepted, go into Production, get Invoiced etc. Basically a funnel. Now it happens that some projects have gone into Quotation multiple times, and I want to retrieve the earliest date that a project went into Quotation in a Calcultated Column.
Given that, which is the better syntax:
Syntax A:
VAR Projectnr = 'Projects'[Projectnumber]
RETURN
MINX (
FILTER (
'Projects',
'Projects'[Projectnumber] = Projectnr
&& 'Projects'[statusnr] = 14
),
'Projects'[Date]
)
Syntax B:
VAR Projectnr = 'Projects'[Projectnumber]
RETURN
CALCULATE (
MIN ( 'Projects'[Date] ),
FILTER (
'Projects',
'Projects'[Projectnumber] = Projectnr
&& 'Projects'[statusnr] = 14
)
)
I wonder if there is a core difference between these two and if so what is the difference in characteristics (speed, higher chance of unexpected results)?
As Syntax A uses CALCULATE I wonder if that induces a less optimal context transition based calculation whereas Syntax B might not use context transition? Or do they both use context transition but is it not implicit in the syntax with Syntax B?
Also, is there a way to measure performance for Calculated Columns in a similar way you can use performance analyzer for report visuals?
