Statalist


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

st: Programs for text processing: a chronology (updated)


From   Kit Baum <baum@bc.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: Programs for text processing: a chronology (updated)
Date   Wed, 3 Feb 2010 21:18:31 -0500

<>
AMENDED 20100203.2115

In re various claims about user-written text processing programs for use with Stata, I would like to present the following facts, in chronological order:

(1) -outreg-, by John Luke Gallup, was written for version 6 of Stata. It was first hosted on the SSC Archive, which I maintain, on 10 Apr 1999. (RePEc templates, unlike package descriptions, contain both creation dates and revision dates). -outreg- was presented in Stata Technical Bulletin 46 (Nov 1998), with revisions in STB 49 (May 1999), STB 58 (Nov 2000) and STB 59 (Jan 2001). The latest version available from SSC is 3.1.8, revised as of 3 May 2008. Given that, I believe that John Gallup would disagree with the statement that appears in -findit outreg- listings for the STB that "The original outreg command is no longer maintained. It has been replaced by outreg2."

-outreg-, being written for Stata 6.0, relies on retrieving information from the ereturn list directly after an estimation command is executed.

(2) The -estimates- suite of commands was added to Stata in version 8 of 2003, permitting a user program to access several sets of estimates from earlier estimation commands.

(3) -estout-, by Ben Jann, was written for version 8.0 of Stata (it now requires Stata 8.2). It was first hosted on the SSC Archive on 22 July 2004. It was subsequently described in Stata Journal 5:3 (Q3 2005) and SJ 7:2 (Q2 2007). It is also documented in a website, http://repec.org/bocode/e/estout/.

-estout- is based on the technology introduced in Stata 8.0: the stored -estimates-. One must use -estimates store- or its own convenience command -eststo- to place stored results where they can be accessed by -estout-. Thus, although the end result is similar to that of -outreg-, the logic is quite different, as -estout- is only invoked after all estimation commands to be tabulated in one table have been executed and their results stored.

The update of -estout- described in the SJ 2007 article addressed the issue that -estout-'s many options made it rather cumbersome to use. The convenience command -esttab-, introduced in this update, provided a 'user-friendly wrapper for the estout command' that used a set of default options to simplify the command (as indicated in the title of the SJ article, "Making regression tables simplified".) At the same time, the -eststo- command (also available as a prefix command) addressed the critique that -estout- required the user to create and manage stored estimates. Although -eststo- and -esttab- are now the preferred way of invoking -estout- to meet many users' needs, it should be clear that they are only convenient alternatives to the explicit use of -estimates store- and -estout-.

(4) -outreg2-, by Roy Wada, was written for version 7 of Stata, and first hosted on the SSC Archive on 19 October 2005 following an announcement on Statalist on 14 September 2005. It has not been documented in the STB or SJ. It is documented as being based to some degree on an uncirculated version 4.0 of -outreg- provided to me by John Gallup. As -outreg2- was written for Stata 7, which did not support stored -estimates-, it originally relied on the same technology as -outreg-: retrieval of results from the ereturn list following an estimation command. It has since been enhanced to work with stored -estimates-, introduced in Stata version 8, but does not require their use. -outreg2- provides additional functionality to users of Stata 8.0, 8.2, and 11.0, as described in its help file. It may be used as a prefix command.


Although this chronology does not (and is not meant to) decisively confirm or reject any claims of offenses against intellectual property rights, I think it is useful to note the order in which these highly popular alternative packages for text processing in Stata were originally developed and made available to the Stata user community.

I do not mean to denigrate the contributions of many other Stata user-programmers (myself included) who have contributed useful tools for specialized text processing (table output, LaTeX output, output to XML, etc.) to the user community. I focus on the three packages mentioned above as they consistently appear on the list of the most frequently downloaded packages in the SSC Archive.

Kit Baum
maintainer, SSC Archive
Research Papers in Economics (RePEc)
http://repec.org




*
*   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