Identify range of macros to be dropped.

[David Speed]

Hi Everyone,

I'm working on a project where I have to define a lot of local variables and will reuse those naming conventions each loop. Given the scope of the project I would like to wipe the locals manually to prevent 'bleed-through' from one loop to the next. However, there are locals that I periodically define before and after that I would like to preserve.

Is there a way to wipe a range of locals?

Apologies for the bolded pseudo-code, but it conveys the basic essence and the desired result in my example. I've gone through the pmacro documentation but I can't find anything quite describing what I'm talking about.

. local a = 1 

. local b = 2 

. local c = 3 

. local d = 4

. local e = 5

. macro list 
_a:             1
_b:             2
_c:             3
_d:             4
_e:             5

macro drop _b - _d

. macro list 
_a:             1
_e:             5

Is there a way to do this? I know there are work arounds to prefixing locals and dropping based on prefixes, but that will be annoying to redefine every local. Alternatively I could use globals to keep the macros I want and drop locals, but I'd prefer to avoid globals if possible.

Thanks for any insight into this!

Cheers,

David.

[Nick Cox]

You seem to have answered your own question. At one extreme, macro drop _all drops every local and you can define the others at will.

[Hemanshu Kumar]

I see no particular reason to define globals. Or redefine locals to have certain prefixes. Why not just keep track of the locals to drop as

local to_drop _b _c _d

and eventually

macro drop `to_drop'

[David Speed]

Hemanshu Kumar

Yes, that will work, but it is also more labour intensive than what I'm seeking. It would require me to redefine a local after each of the 100 locals I am already defining. This is why I was looking for a function that worked similar to how ranges of variables can be dropped. To clarify, I can absolutely brute force the solution, but 95% of the time Stata has a cool feature that I'm simply unaware of.