whatsnew6.hlp

是一个经济学管理应用软件 很难找的 但是经济学学生又必须用到

{p 0 4}
{help mhodds} now displays better headers when option {cmd:by()} is specified
with string variables.

{p 0 4}
{help streset} would fail if the {cmd:before} option was specified.

{p 0 4}
{help xtnbreg} now honors the {cmd:nolog} option.

{hline 5} {hi:update 28jun1999} {hline}

{p 0 4}
{help arima} sometimes produced unclear and misleading messages during
optimization.  Messages have been reworded.

{p 0 4}
{help cf}'s option {cmd:verbose} did nothing; it is fixed.

{p 0 4}
{help collapse} no longer retains the name of the original dataset and so
typing "{cmd:save, replace}" after {cmd:collapse} will not replace the
original, uncollapsed data.

{p 0 4}
{help compare} run on some files that contained string variables presented a
syntax error and stopped; that is fixed.

{p 0 4}
{help corrgram}, {help ac}, and {help pac} have been given minor improvements:
{cmd:corrgram} now stores results in {hi:r()}, {cmd:ac} is faster, and
{cmd:pac} handles errors better.

{p 0 4}
{help fracpoly}.  Setting the random-number seed did not recreate an analysis
when the {cmd:add} option was specified.

{p 0 4}
{help lsens}'s {cmd:replace} option did not work; it now does.

{p 0 4}
{help stsplit} and {help stjoin}.  {cmd:stsplit}'s way of recording time could
cause {cmd:stjoin}, in some cases, not to be able to rejoin all the records
possible.

{p 0 4}
{help svytab} with the {cmd:subpop()} option mistakenly printed the word
"dash" in its header; it no longer does.

{hline 5} {hi:update 22jun1999} {hline}

{p 0 4}
{hi:Stata executable}(*), all platforms.{p_end}
{p 4 4}
HTTP proxy authorization has been added.

{p 4 4}
In processing numlists, Stata issued an error when numlists of the form
"{it:#}{cmd:(}{it:#}{cmd:)}{it:#}" or "{it:# #} {cmd:to} {it:#}" did not start
with an integer element.

{p 4 4}
The {help numlist} parsing command issued an out-of-range error when option
{cmd:range(<}{it:#}{cmd:)} was specified even when the list did not contain
out-of-range elements.

{p 4 4}
{help summarize} now saves {hi:r(sd)} = sqrt({hi:r(Var)}).

{p 4 4}
{help tabulate} truncated a character in the labels of one-way tabulations
when the labels were longer than 8 or 9 characters.


{p 0 4}
{hi:Stata for Windows executable}(*).{p_end}
{p 4 4}
{hi:--more--} conditions now work as expected with dialog boxes.

{p 4 4}
The do-file editor now clears the edited filename when a new instance is
started.  Thus, restarting the do-file editor and clicking the save button now
results in the do-file editor asking for the name under which the file is to
be saved.

{hline 5} {hi:update 31may1999} {hline}

{p 0 4}
On-line help and search index brought up to date for STB-49.

{p 0 4}
{help cloglog} with the {cmd:noconstant} option sometimes dropped variables
when it should not have.  That has been fixed and this updated version also
converges more reliably.

{p 0 4}
{help lrtest} now handles models with no RHS variables.

{p 0 4}
{help ml} {cmd:graph} did not work when there were more than 10 iterations; it
now does.

{p 0 4}
{help xtgls} would, for some models, reset the {cmd:iis} and {cmd:tis}
variables so that they were interchanged.  This is fixed.

{hline 5} {hi:update 13may1999} {hline}

{p 0 4}
{help collapse} previously produced a {help float} result when working with
{help long} variables and now produces a {help double} result, thus
maintaining all possible precision.

{p 0 4}
{help heckprob} now works correctly for selection dependent variables that are
zero/not zero.  Previously, it required that the selection dependent variable
be 0/1.

{p 0 4}
{help reshape} mislabeled the heading of its report when reshaping from long
to wide.  It said "wide -> long" when it should have said "long -> wide".

{p 0 4}
{help stset}'s {cmd:exit(}{it:var}{cmd:=}{it:numlist}{cmd:)} option would not
work if {help numlist} contained more than one variable.

{hline 5} {hi:update 05may1999} {hline}

{p 0 4}
{help prais} issued an error message when used with more than 10 or 20
variables.  Now it is fixed.

{hline 5} {hi:update 27apr1999} {hline}

{p 0 4}
{help arima} and {help arch}.  To make the interpretation of dynamic
predictions more consistent for a wider range of models, the meaning of the
{cmd:dynamic()} option has changed.  Now, all references to the dependent
variable, say y, prior to the time period specified in {cmd:dynamic()}
evaluate to the observed (not predicted) value of y; all values of y on or
after {cmd:dynamic()} evaluate to the predicted value of y.

{p 0 4}
{help xi} would not work with time-series operators if the dataset was
{help tsset} as panel data.

{hline 5} {hi:update 26apr1999} {hline}

{p 0 4}
{help tabi} now works correctly with larger than 9 x k or k x 9 tables.

{hline 5} {hi:update 23apr1999} {hline}

{p 0 4}
{help grmeanby} did not work when weights were specified.

{p 0 4}
{help ivreg} would fail if the number of instruments exceeded about 450.

{p 0 4}
{help ltable} did not present the log-rank test if weights were specified.

{p 0 4}
{help pause} now runs in version-6 mode, meaning commands you type in debug
mode are interpreted according to the modern rules.

{p 0 4}
{help sttocc} did not necessarily reselect the same "random" sample when rerun
with the same random-number seed; it now does.  In addition, {cmd:sttocc}'s
{cmd:match()} option now allows a varlist rather than just a varname.

{p 0 4}
{help stmh} did not, under certain conditions, report results when option
{cmd:by()} was specified.

{p 0 4}
{help sts} now uses the value labels to label graphs when you have attached
value labels to the analytic time variable {hi:_t} (as unlikely as that may
be).

{hline 5} {hi:update 21apr1999} {hline}

{p 0 4}
{hi:Stata executable}(*), all platforms.{p_end}
{p 4 4}
{cmd:cox} and {help stcox} now have a new {cmd:noadjust} option for use with
the {cmd:robust} option.  This option prevents the estimated variance matrix
from being multiplied by N/(N-1) or g/(g-1) (g the number of clusters).  Most
users will probably not wish to use this new option.

{p 4 4}
{help gettoken} and {help tokenize} now support long macros.

{p 4 4}
*.pkg files may now contain "{hi:f ../}etc." references to mean parent
directory.

{p 0 4}
{hi:Stata executable}(*), Unix.{p_end}
{p 4 4}
Some users saw $<#> in help files when text was supposed to be highlighted.
The text is now shown in highlighted form just as intended.

{hline 5} {hi:update 12apr1999} {hline}

{p 0 4}
On-line help and search index brought up to date for STB-48.

{hline 5} {hi:update 02apr1999} {hline}

{p 0 4}
{help adjust}.  The reported standard errors were incorrect when some
variables were left unspecified.  Results were correct when every variable in
the model was specified either in the varlist or in the {cmd:by()} option.

{p 0 4}
{help arima} and {help arch} could not estimate models if there were
insufficient contiguous observations for the starting value algorithm.  In
such cases, they now default to starting values of 0 for the ARMA and/or ARCH
parameters.

{p 0 4}
{help hausman} has new options {cmd:prior()} and {cmd:current()} to allow the
labels for the coefficient columns to be changed.

{p 0 4}
{help heckman}.  The counts of censored and uncensored observations were
reversed in the output.

{p 0 4}
{help heckprob}.  (1) The {cmd:noconstant} option was not honored in the
selection equation; it now is.  (2) When there were missing values among the
RHS variables of the primary equation for observations that were not selected,
these observations were incorrectly omitted from the estimation sample.  (3)
Better starting values are now used so models should converge more quickly.

{p 0 4}
{help integ} incorrectly required the user to specify the {cmd:generate()}
option.

{p 0 4}
{help istdize} would sometimes mistakenly issue the error message that the
number of cases exceeded the population size and so refused to run, even when
that was not true.

{p 0 4}
{help kdensity}'s {cmd:title()} option now understands {cmd:title(" ")} to
mean no title is to be displayed and the {cmd:title()} option will accept
compound double quotes.

{p 0 4}
{help quadchk} now works with {help xtintreg}.

{p 0 4}
{help scobit}'s model test is deleted because the ancillary parameter alpha
and the constant are not both identifiable in the constant-only model.
{cmd:scobit} now also converges more reliably.

{p 0 4}
{help xtgls}.  (1) {cmd:xtgls} now reports the correct log-likelihood.  (2)
Previously, if option {cmd:igls} was specified, {cmd:xtgls} reported
consistent, iterated results, but these were not the MLE; now MLE results are
produced.  (3) Previously, when {cmd:panel(correlated)} was specified without
option {cmd:igls}, the GLS matrix was based on a panel-heteroskedastic model.
It is now computed from OLS.  Both yield consistent results but the new
approach is more in line with what most people would expect.

{p 0 4}
{help xtnbreg}{cmd:, fe} now explicitly omits all observations that do not
contribute to the likelihood (e.g., groups of size one).  {cmd:xtnbreg}, both
{cmd:fe} and {cmd:re}, also now converges more reliably for difficult problems.

{p 0 4}
{help xtpois}{cmd:, fe} now explicitly omits all observations that do not
contribute to the likelihood (e.g., groups of size one).
{cmd:xtpois, re normal} now converges more reliably for the constant-only
model.

{hline 5} {hi:update 19mar1999} {hline}

{p 0 4}
{help snapspan} refused to transform the data and instead issued a syntax
error when there was more than one event variable in the dataset.

{p 0 4}
{help tabodds}.  When weights were used, the score test for trend reported by
{cmd:tabodds} was incorrect.  (All other statistics were fine.)  The problem
has been fixed.

{hline 5} {hi:update 15mar1999} {hline}

{p 0 4}
{help nbreg} has been enhanced to optionally estimate an alternative
parameterization of the negative binomial model.  Choosing the option
{cmd:dispersion(constant)} yields a model in which the overdispersion is
constant from observation to observation.  See help {help nbreg} for details.

{p 0 4}
{help stphplot} would sometimes connect the points in the resulting graph in
an odd order.

{p 0 4}
{help xtnbreg}{cmd:, fe} should have included a constant term, but it did not.
When weights were specified in the {cmd:re} and {cmd:fe} models, it would not
converge to the right answer; it now converges reliably.  The comparison model
for the {cmd:re} case is the new {cmd:nbreg, dispersion(constant)} alternative
parameterization.  The command is now much faster as well.

{p 0 4}
{help xtpois}{cmd:, re} (without weights) could sometimes have a problem
converging.  When weights were specified in the {cmd:re} and {cmd:fe} models,
it would not converge to the right answer.  It now converges reliably with and
without weights, and it is much faster as well.

{hline 5} {hi:update 04mar1999} {hline}

{p 0 4}
{help fracpoly}.  The way the {cmd:adjust()} option works has changed.

{p 0 4}
{help sw} now works with {help ologit} and {help oprobit}.

{p 0 4}
{hi:Stata executable}(*), all platforms.{p_end}
{p 4 4}
When version is set to 5, {help label} {cmd:save} now produces files
compatible with Stata 5.0.

{p 4 4}
{cmd:cox} and {help stcox} no longer allow specifying option {cmd:robust}
with {cmd:mexact} or {cmd:pexact} options.

{p 4 4}
{help ologit} and {help oprobit} now report the Wald model chi2 statistic if
option {cmd:offset()} is specified.

{p 4 4}
{help inspect} caused {help fracpoly} to crash under certain circumstances;
that is fixed.

{p 0 4}
{hi:Stata for Macintosh executable}(*).{p_end}
{p 4 4}
profile.do can now be located in a f