Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: In defense of Stata, was RE: syntax question

From   "David Moore" <>
Subject   st: In defense of Stata, was RE: syntax question
Date   Tue, 19 Nov 2002 13:55:47 -0800

Although I complain about the lack of a facility for user defined functions,
there is one area in which Stata's macro language surpasses SAS's.  Due to
the way SAS implements macros, it is very cumbersome for a SAS program (the
rough equivalent of a Stata "do" file) to communicate with macros.  In
Stata, you can write statements, such as "local num_obs = _N", to treat the
value of _N as text in subsequent macro program code.  For example, I can
use the macro variable just created in lines, such as "gen var`num_obs' =
_n."  You can't easily do this in SAS.  In effect, Stata makes no
distinction at run-time between program code in an "ado" and code in a "do"
file that calls the "ado" program, whereas SAS translates all user written
procedures and functions into a standard SAS program prior to run-time.

Most important, Stata has built-in support for programmers interested in
extending it with procedures that resemble OEM commands.  The error checking
for commands that follow standard syntax is a significant time saver for
those of us who want to maintain the look and feel of Stata when writing our
own procedures.

-----Original Message-----
[]On Behalf Of Nick Cox
Sent: Tuesday, November 19, 2002 12:50 PM
Subject: st: RE: RE: RE: RE: RE: syntax question

David Moore

> There's at least one thing you can do with the SAS macro programming
> language that I have found useful.  You can create
> functions.  I know egen
> allows some limited capabilities with respect to new
> variable generation,
> but functions are more general.  For example, they can be
> used to calculate
> or return values in many different programming contexts,
> not just to create
> variables.

True. That is, suppose you wanted to write
a Stata function to calculate logits. You could write
a -egen- function to create new variables, or
you could write -logitf- such that you could go

logitf 0.5

and it could print 0, and leave behind


so that you could go

logitf <argument>
... = r(logit)

So you can't define a function allowing you then to do
things like

. di logitf(0.5)

in just one line.


*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index