Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Is -_ms_parse_parts- here to stay?


From   vwiggins@stata.com (Vince Wiggins, StataCorp)
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Is -_ms_parse_parts- here to stay?
Date   Thu, 06 Aug 2009 17:22:50 -0500

Roger Newson <r.newson@imperial.ac.uk> asks,

> [...] Is the undocumented -_ms_parse_parts- utility here to stay?
> (And are the other undocumented -_ms_- utilities also here to stay?)
> I ask because I would like to use this utility (or something
> similar) in my own Stata 11 packages, if it is not likely to
> disappear without warning in the near future. In particular, I would
> like to regenerate factor variables in -parmest- resultssets, the
> way I used to do with the SSC package -factext- (Newson, 2004). [...]


Short Answer
------------

I recommend Roger use -_ms_parse_parts- and the other -_ms_- utilities freely
in his programs.  I say that for two reasons. One, I do not think they will be
removed or their current syntax made to fail in future releases.  Two, if he
wants to handle factor variables in his programs, these utilities will be
necessary.


Long Answer
-----------

This has been mentioned before on the list, but let me recapitulate what
"undocumented" means.  If you type in Stata -help _ms_parse_parts- you will
get a help file!  What's more, at the upper right of that DOCUMENTATION is a
blue link that will take you to an index of other "undocumented" commands!
Those "undocumented" commands are also documented in help files.  There are a
very few other commands that have no documentation that we believe would not
be of use even to user programmers and those commands are "not documented".
It is best not to think about it too hard.

There is a nice description of what "undocumented" means at the top of -help
undocumented- and I will leave the interested to read that.  These commands
are "undocumented" primarily because we believe they are of limited interest,
though they might also be a work in progress.

One thing that is different about "undocumented" commands is that StataCorp
reserves the right to change them in the future in such a way that their
current syntax will no longer work, even under version control -- something we
have almost never done with official commands or any aspect of Stata syntax.
That said, I can't recall us making such a change in the 3 releases we have
had "undocumented" commands.

So, I am comfortable recommending that Roger use the "undocumented" commands
for handling factor-variable varlists.  I would not be surprised if we add
some documented commands in the future that duplicate some of the features of
these "undocumented" commands.  If we do, it is very likely that the
"undocumented" commands will continue to work as they do now, even though they
might no longer be the preferred solutions.


Roger goes on to ask,

> And are there any other useful utilities that would be useful for
> parsing the column stripe elements of -e(b)-, and which are not
> documented, even under -help undocumented-?

No.  We have "undocumented" everything we thought useful.


-- Vince 
   vwiggins@stata.com

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



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