[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: Re: estout vs outreg outreg2
There have been some discussion about outreg and outreg2. Perhaps I should
have explained it more clearly when I first released outreg2.
First, outreg2 is not a wrapper and does the heavy lifting itself, but the
task has been made easier by the good people of the Stata Corp, who had
established a stable nomenclature for matrix names. This means outreg2 can
handle ALL estimation commands that conforms to the official guideline.
Hence outreg2 is extremely stable and needs very little updating. It goes
without saying (until now) that the most obvious functions, such as the
recently raised issue of directly reporting the raw t-statistics, has been
implemented within outreg2. All idiosyncratic issues (known to me) have been
Second, outreg2 will handle a list of STORED ESTIMATES made by -estimates-
suite of Stata command. For example, to report your stored estimates named
TOM and JERRY, or all of the stored estimates, do something like this:
. outreg2 [TOM JERRY] using myfile
. outreg2 [*] using myfile
Third, outreg2 will produce tables in Word, Excel, LaTeX, and text format.
outreg2 also comes with -shellout- command that DIRECTLY OPENS the produced
tables in Word, Excel, or LaTeX (or TeX-associated editors) for the users of
Window XP/NT. This is very fast and a useful tool during research. A blue
hypertext is also produced, which you can click to activate.
Fourth, outreg2 comes with -seeout- command that will immediately display
the produced table in the BROWSER VIEW WITHIN Stata. This is also fast
(unless you have a million-plus observations). A blue hypertext is produced
that you can click.
Philosophically speaking, outreg2 is a research tool to be used DURING
research, not AFTER. Making regression tables is not something you would
perform only at the end of research. Without undue effort on your part,
outreg2 gives you the immediate access to a formatted table that let you
compare ACROSS regressions on the computer screen. SAS users have to
PHYSICALLY PRINT their outputs in order to compare them MANUALLY. This is
why the SAS users print hundreds of pages. (A note to system administrators:
you can realize a significant cost reductions in paper/toner by switching to
I talk about these issues (among other things) in a paper that I had written
last year. I probably should have submitted it to the Stata Journal, but
outreg2 does not really need an extensive documentation to use it. The ease
of use is a virtue when you are worrying about something more important
(like choosing a model). I am not into having to look up a command everytime
I use it. I often find myself using a workaround just to avoid having to
look it up.
P.S. I suppose I could add several lines to outreg2 to make it fully
compatible with estadd (okay with scalars, but not with matrices); what I
would like to see as a user is a program that would handle summary
statistics the way outreg2 handles the regression outputs. I suppose it can
be done, using the ereturn matrices or by-passing them.
First of all the relevant comparison is estout versus outreg2. outreg has
not been developed for some time, which triggered Roy Wada's development of
The primary difference is that outreg2 (like outreg) does all the heavy
lifting itself, whereas estout piggybacks upon the official - estimates-
suite of Stata commands. Thus additional tools such as estadd would not be
useful in conjunction with outreg2, since outreg2 does not make use of
I used to use outreg, and tested outreg2 when it appeared. estout is
complex, but also is very fully documented by an extensive article in the
Stata Journal. estout also has a useful "intro" help file which shows how
it can be used without any complexity, taking the defaults for many of the
options. I highly recommend that anyone interested in output to text, HTML
or LaTeX have a look at estout.
Kit Baum, Boston College Economics
* For searches and help try: