SPSS Command to STATA

[Paige LaPierre]

Hello all;

Anybody know how to transform the following SPSS command into a STATA command?
Thank you

USE ALL.
COMPUTE filter_$=(B8_under5 = 1 | B8_under5 = 2 | B8_under5 = 3).
VARIABLE LABELS filter_$ 'B8_under5 = 1 | B8_under5 = 2 | B8_under5 = 3 (FILTER)'.
VALUE LABELS filter_$ 0 'Not Selected' 1 'Selected'.
FORMATS filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE.

[Nick Cox]

I haven't used SPSS in the current millennium but I would guess that a translation to Stata (*) starts

gen wanted = inlist(B8_under5, 1, 2, 3) 
label var wanted "B8_under5 1, 2 or 3" 
label def wanted 0 "Not selected" 1 "Selected" 
label val wanted wanted

where you should choose your own variable name.

(*) Not STATA. https://www.statalist.org/forums/help#spelling

[Mike Lacy]

To Nick's correct code, I would add the note that the -filter by- command in SPSS does not fit with a typically Stata-ish way of doing things. To my recollection (?), SPSS's -filter by- command selects for analysis cases with the filter variable = 1, while excluding others. This restriction stays in effect until filtering is turned off. Analogous behavior can be produced in Stata with a preserve/restore sequence:

preserve
keep if wanted == 1
.... do whatever analysis is desired ...
restore

However, in most situations, things one might do by "filtering" in SPSS are accomplished in Stata with the if "qualifier" construction, without any preserve/restore construction, of which a prototype would be:

DoMyAnalysis if wanted == 1

So, while the code can be literally translated as Nick shows, a more idiomatic translation might do good things for Paige.