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

Re: st: Stata 9 announcement

From   Ulrich Kohler <>
Subject   Re: st: Stata 9 announcement
Date   Mon, 4 Apr 2005 13:30:04 +0200

I take the oportunity to remind all of you, that the German Stata Users' Group 
Meeting will be on Friday. Bill Gould and David Drukker will be on the 
meeting, and will obviously talk about Stata 9. The meeting will be therefore 
a perfect place to get the latest News about Stata 9.

For Details on the meeting, please check out

many regards,

William Gould, Stata wrote:
> We have a tradition of telling Statalist members about new releases of
> Stata before anyone else, so here it is:
> I am pleased to be the one to announce the release of Stata 9.
> Stata 9 begins shipping on April 27th.
> Highlights include
>     o  New matrix language Mata, with 620-page manual.
>        Mata is a full-blown matrix programming language that compiles
>        what you type into byte-code, optimizes it, and executes it fast.
>        You can use Mata interactively, or you can use Mata to implement
>        big systems.   We used Mata to implement many of the new features
>        of Stata 9.
>     o  New survey features, including balanced repeated replications
>        (BRR) and jackknife variance estimates, multistage designs, and
>        poststratification.
>     o  Linear mixed models, including two-way, multilevel, and
>        hierarchical random-effects -- random intercepts and
>        random coefficients.
>     o  Multinomial probit models, including support for several
>        correlation structures and support for user-defined correlation
>        structures.
>     o  Probit and tobit with endogenous regressors, including
>        maximum-likelihood and two-step estimators.
>     o  Stereotype logistic regression (relaxes the assumption of
>        proportional odds).
>     o  Zero-truncated Poisson and negative binomial.
>     o  Multivariate analysis, including multidimensional scaling,
>        correspondence analysis, biplots, tetrachoric correlations, and
>        Procrustean analysis, along with the ability to analyze proximity
>        matrices and raw data.
>     o  Seasonal ARIMA models, including robust standard errors.
>     o  Nonlinear least squares that is now as easy to use as linear
>        least squares, and including clustered-robust standard errors.
>     o  Bootstrapping and jackknifing that is easier to use, including
>        full VCE estimates.  (You can now just add -vce(bootstrap)- or
>        -vce(jackknife)- to your estimation command.)
>     o  XML support.
>     o  Enhanced GUI, including multiple Do-file editors, multiple
>        Viewers, multiple Graph windows, all of which can move outside
>        the main Stata window; multiple windowing preferences; dockable
>        windows; copy text to HTML; etc.
>     o  More 64-bit support, including 64-bit Windows and 64-bit Linux on
>        Itanium, Opteron, Athlon 64, and EM64T.  (Still available are
>        AIX, IRIX, Solaris, and Tru64.)
> I've just given the highlights.  Other features include rolling and
> recursive regressions; robust and clustered-robust SEs for random-
> effects and fixed-effects regression; mean, ratio, total, and proportion
> estimators for groups with full VCE; paired-coordinate plot types;
> value labels up to 32,000 characters, etc.
> For more information, visit  You can order
> from there, or you can send email to
> Just do not send email directly to me --  I'm heading
> off to the German users meeting and will not see your email until I
> return.
> -- Bill
> P.S.  Every time I announce a new release, a few people are concerned
>       that, having just bought Stata, they need to buy again.  Not true.
>       There is special upgrade pricing for those who purchased recently.
>       The rule is simple:  if you purchased in January, February, March,
>       or April, you get the upgrade for the difference in cost, and
>       there is a discount for those who purchased in December.
>       In these special cases, the easiest way to order is to call
>       1-800-STATAPC or send email to
>       Of course, you can always call or send email if you have
>       questions.
> <end>
> *
> *   For searches and help try:
> *
> *
> *

+49 (030) 25491-361

*   For searches and help try:

© Copyright 1996–2022 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index