Question
· 2 hr ago

Conditions quantity in $SELECT()

#Analytics #Analyzer #InterSystems IRIS #InterSystems IRIS BI (DeepSee)

I use $SELECT() in Architect to aggregate various data conditions into a single dimension.

I have a data table like this: id, user, userData1, userData2...

Where each user action is the %Integer ID of that action from some source. Then in expression field i use: $SELECT(userAction1:"This user action", userAction2:"That user action"...) and so on, up to 14 conditions, excluding the default placeholder 1:"Other" (which is always empty because there is a condition with 100% chance). The logic is simple: if userAction1 is not null then SELECT return "This user action" and so on. The end result is that I get a perfect measurement with all the user data labeled the way I want. (Until the sum of the conditions hits an invisible wall)

After that, if we check the dimension in the analyzer, we find that only 10 conditions fall into this dimension; the rest are missing somewhere

I checked the documentation, and there's no description of the hard limit. I asked dc ai and he said the same thing. Perhaps the hard limit is an undocumented parameter?

Product version: IRIS 2025.3
Discussion (2)2
Log in or sign up to continue
Robert Cemper · 29 min ago

I composed a small example as DV.MAC  with  36 conditions. and  -as expected- hit no limit.
 

DV ;
	for i=1:1:50 {
		set x=$r(40)
		set col=$SELECT(
 			x=1:"col_1",
			x=2:"col_2",
			x=3:"col_3",
			x=4:"col_4",
			x=5:"col_5",
			x=6:"col_6",
			x=7:"col_7",
			x=8:"col_8",
			x=9:"col_9",
			x=10:"col_10",
			x=11:"col_11",
			x=12:"col_12",
			x=13:"col_13",
			x=14:"col_14",
			x=15:"col_15",
			x=16:"col_16",
			x=17:"col_17",
			x=18:"col_18",
			x=19:"col_19",
			x=20:"col_20",
			x=21:"col_21",
			x=22:"col_22",
			x=23:"col_23",
			x=24:"col_24",
			x=25:"col_25",
			x=26:"col_26",
			x=27:"col_27",
			x=28:"col_28",
			x=29:"col_29",
			x=30:"col_30",
			x=31:"col_31",
			x=32:"col_32",
			x=33:"col_33",
			x=34:"col_34",
			x=35:"col_35",
			1:"more")
		write i,?4,x,?10,col,!
	}

result:

. . . . 
23  32    col_32
24  39    more
25  39    more
26  31    col_31
27  10    col_10
28  36    more
29  24    col_24
30  21    col_21
31  2     col_2
. . . .
0 0