Statalist The Stata Listserver

[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: How to make macro variables last an entire session?

From   "Nick Cox" <>
To   <>
Subject   st: RE: How to make macro variables last an entire session?
Date   Tue, 29 Aug 2006 13:27:04 +0100

There is some confusion here between _where_ and _when_. 

A local macro (don't attach the word "variable" here, 
as it doesn't apply) can last a whole session so
long as it is defined interactively and you 
don't overwrite or delete it. Thus if you were 
to define your -local base ...- in the Command 
window this would stick around, so long as you 
yourself don't zap it, to the end of the session. 

By defining a -local- within a .do file you 
are saying to Stata that the macro should persist
only to the end of the file (which here by 
courtesy is considered a kind of program). 

That's what "local" means, precisely: _local_ to the 
program in which it is defined, where "program" 
is in a wide sense one of 
	a sequence of commands typed interactively 
	a -program- explicitly defined (e.g. within
		an .ado file)
	a .do file 
	the contents of the do-file editor executed. 

There is a way of defining locals that will persist 
from within .do files. It is called -c_local- and
it is not documented (it is not even under -undocumented-).
The usual coy advice is that you shouldn't use it until 
you have worked out for yourself from examples how it 
works and why it can be dangerous. 

All that said, the advice of Philipp Rehm to use -global-s
instead is good. But globals also will only last an 
entire session if they are not zapped. 

> I was wondering if anyone knows how to define or save macro 
> variables so they last for an entire stata session and not 
> just while the do-file is running. I've been defining macro 
> variables  as is below with the local command where the 
> command `base' substitutes for all the variables in quotes 
> below for example. But if I run the local command, and then 
> later highlight some line of code with `base' and "do" it, 
> then Stata no longer associates `base' with the set of 
> variables. Anyone know any ways around this problem? I've 
> tried saving using global macros and that doesn't seem to help.
> local base "lncgdp s n popuned popprimed popseced pophighed openc"
> regress `base'

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index