whatsnew7to8.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 6 10 2}
8.  The existing {cmd:arima} command may now be used with the {cmd:by} prefix
    command, and it now allows prediction in loops over panels; see help
    {help arima}.

{p 6 10 2}
9.  The existing {cmd:newey} command now allows (and requires) that you
    {help tsset} your data; see help {help newey}.


{title:What's new in cross-sectional time-series analysis}

{p 6 10 2}
1.  The new {cmd:xthtaylor} command fits panel-data random-effects models
    using the Hausman-Taylor and the Amemiya-MaCurdy
    instrumental-variables estimators; see help {help xthtaylor}.

{p 6 10 2}
2.  The new {cmd:xtfrontier} command fits stochastic production or cost
    frontier models for panel data allowing two different parameterizations
    for the inefficiency term: a time-invariant model and the Battese-Coelli
    (1992) parameterization of time effects; see help {help xtfrontier}.

{p 6 10 2}
3.  The existing {cmd:xtabond} command now allows endogenous regressors; see
    help {help xtabond}.

{p 6 10 2}
4.  The existing {cmd:xtivreg} command will now optionally report first stage
    results of Baltagi's EC2SLS random-effects estimator; see help
    {help xtivreg}.

{p 6 10 2}
5.  The existing {cmd:xttobit} and {cmd:xtintreg} commands have new
    {cmd:predict} options:

{p 10 14 2}
    a.  {cmd:pr0(}{it:#_a}{cmd:,}{it:#_b}{cmd:)} produces the probability of
	the dependent variable being uncensored P({it:#_a}< y < {it:#_b}).

{p 10 14 2}
    b.  {cmd:e0(}{it:#_a}{cmd:,}{it:#_b}{cmd:)} produces the corresponding
	expected value E(y | {it:#_a} < y < {it:#_b}).

{p 10 14 2}
    c.  {cmd:ystar(}{it:#_a}{cmd:,}{it:#_b}{cmd:)} produces the expected
	value of the dependent variable truncated at the censoring point(s),
	E(y^*), where y^* = max({it:#_a}, min(y,{it:#_b})).

{p 10 10 2}
  See help {help xttobit} and {help xtintreg}.

{p 6 10 2}
6.  Existing commands {cmd:xtgee} and {cmd:xtlogit} have a new {cmd:nodisplay}
    option that suppresses the header and table of coefficients; {cmd:xtregar,
    fe} now allows {cmd:aweight}s and {cmd:fweight}s; and {cmd:xtpcse} now has
    no restrictions on how {cmd:aweight}s are applied.  See help {help xtgee},
    {help xtlogit}, and {help xtpcse}.

{p 6 10 2}
7.  Two commands have been renamed:  {cmd:xtpois} is now called
    {cmd:xtpoisson} and {cmd:xtclog} is now {cmd:xtcloglog}.  The old names
    continue to work.  See help {help xtpoisson} and {help xtcloglog}.


{title:What's new in survival analysis}

{p 6 10 2}
1.  Existing command {cmd:stcox} has an important new feature and some minor
    improvements:

{p 10 14 2}
    a.  {cmd:stcox} will now fit models with gamma-distributed frailty.  In
	this model, frailty is assumed to be shared across groups of
	observations.  Previously, if one wanted to analyze multivariate
	survival data using the Cox model, one would fit a standard model and
	account for the correlation within groups by adjusting the standard
	errors for clustering.  Now, one may directly model the correlation by
	assuming a latent gamma-distributed random effect or frailty;
	observations within group are correlated because they share the same
	frailty.  Estimation is via penalized likelihood.  An estimate of the
	frailty variance is available and group-level frailty estimates can be
	retrieved.

{p 10 14 2}
    b.  {help fracpoly}, {help sw}, and {help linktest} now work after
	{cmd:stcox}.

{p 10 10 2}
    See help {help stcox}.

{p 6 10 2}
2.  Existing command {cmd:streg} has an important new feature and some minor
    improvements:

{p 10 14 2}
    a.  {cmd:streg} has new option {cmd:shared(}{it:varname}{cmd:)} for
	fitting parametric shared frailty models, analogous to random effects
	models for panel data.  {cmd:streg} could, and still can, fit frailty
	models where the frailties are assumed to be randomly distributed at
	the observation level.

{p 10 14 2}
    b.  {help fracpoly}, {help sw}, and {help linktest} now work after
	{cmd:streg}.

{p 10 14 2}
    c.  {cmd:streg} has four other new options: {cmd:noconstant},
	{cmd:offset()}, {cmd:noheader}, and {cmd:nolrtest}.

{p 10 10 2}
    See help {help streg}.

{p 6 10 2}
3.  {cmd:predict} after {cmd:streg, frailty()} has two new options:

{p 10 14 2}
    a.  {cmd:alpha1} generates predictions conditional on a frailty equal to
	1.

{p 10 14 2}
    b.  {cmd:unconditional} generates predictions that are "averaged" over
	the frailty distribution.

{p 10 10 2}
    These new options may also be used with {cmd:stcurve}.  See help
    {help streg}.

{p 6 10 2}
4.  {cmd:sts graph} and {cmd:stcurve} (after {cmd:stcox}) can now plot
    estimated hazard functions, which are calculated as weighted kernel
    smooths of the estimated hazard contributions; see help {help sts}.

{p 6 10 2}
5.  {cmd:streg, dist(gamma)} is now faster and more accurate.  In addition,
    you can now predict mean time after gamma; see help {help streg}.

{p 6 10 2}
6.  Old commands {cmd:ereg}, {cmd:ereghet}, {cmd:llogistic},
    {cmd:llogistichet}, {cmd:gamma}, {cmd:gammahet}, {cmd:weibull},
    {cmd:weibullhet}, {cmd:lnormal}, {cmd:lnormalhet}, {cmd:gompertz},
    {cmd:gompertzhet} are deprecated (they continue to work) in favor of
    {cmd:streg}.  Old command {cmd:cox} is now deprecated (it continues to
    work) in favor of {cmd:stcox}.  See help {help streg} and {help stcox}.


{title:What's new in survey analysis}

{p 6 10 2}
1.  Stata's {cmd:ml} user-programmable likelihood-estimation routine has new
    options that automatically handle the production of survey estimators,
    including stratification and estimation on a subpopulation; see help
    {help ml}.

{p 6 10 2}
2.  Four new survey estimation commands are available:

{p 10 14 2}
    a.  {cmd:svynbreg} for negative-binomial regression; see help
	{help svynbreg}.

{p 10 14 2}
    b.  {cmd:svygnbreg} for generalized negative-binomial regression; see help
	{help svygnbreg}.

{p 10 14 2}
    c.  {cmd:svyheckman} for the Heckman selection model; see help
	{help svyheckman}.

{p 10 14 2}
    d.  {cmd:svyheckprob} for probit regression with selection; see help
	{help svyheckprob}.

{p 6 10 2}
3.  Use of the survey commands has been made more consistent.

{p 10 14 2}
    a.  {cmd:svyset} has new syntax.  Before it was

{p 18 22 2}
	    {cmd:svyset} {it:thing_to_set} [{cmd:, clear} ]

{p 14 14 2}
	and now it is

{p 18 22 2}
	    {cmd:svyset} [{it:weight}] [{cmd:, strata(}{it:varname}{cmd:)}
		{cmd:psu(}{it:varname}{cmd:)} {cmd:fpc(}{it:varname}{cmd:)} ]

{p 14 14 2}
	See help {help svyset} for details.  In addition, you must now
	{cmd:svyset} your data prior to using the survey commands; no longer
	can you set the data via options to the other survey commands.

{p 10 14 2}
    b.  Two survey estimation commands have been renamed:  {cmd:svyreg} to
	{cmd:svyregress} and {cmd:svypois} to {cmd:svypoisson}; see help
	{help svyregress} and {help svypois}.

{p 10 14 2}
    c.  {cmd:svyintreg} now applies constraints in the same manner as all
	other estimation commands; see help {help svyintreg}.

{p 10 14 2}
    d.  {cmd:lincom} now works after all {cmd:svy} estimators; see help
	{help lincom}.  ({cmd:svylc} is now deprecated.)

{p 10 14 2}
    e.  {cmd:testnl} now works after all {cmd:svy} estimators; see help
	{help testnl}.

{p 10 14 2}
    f.  {cmd:testparm} now works after all {cmd:svy} estimators; see help
	{help test}.

{p 10 14 2}
    g.  The new {cmd:nlcom} and {cmd:predictnl} commands, which form nonlinear
	combinations of estimators and generalized predictions, work after all
	{cmd:svy} estimators; see help {help nlcom} and {help predictnl}.

{p 6 10 2}
4.  Existing command {cmd:svytab} has three new options: {cmd:cellwidth()},
    {cmd:csepwidth()}, and {cmd:stubwidth()}; they specify the widths of table
    elements in the output.  See help {help svytab}.


{title:What's new in cluster analysis}

{p 6 10 2}
1.  The new {cmd:cluster wardslinkage} command provides Ward's linkage
    hierarchical clustering and can produce Ward's method, also known as
    minimum-variance clustering.  See help {help clward}.

{p 6 10 2}
2.  The new {cmd:cluster waveragelinkage} command provides weighted-average
    linkage hierarchical clustering to accompany the previously available
    average linkage clustering.  See help {help clwav}.

{p 6 10 2}
3.  The new {cmd:cluster centroidlinkage} command provides centroid linkage
    hierarchical clustering.  This differs from the previously available
    {cmd:cluster averagelinkage} in that it combines groups based on the
    average of the distances between observations of the two groups to be
    combined.  See help {help clcent}.

{p 6 10 2}
4.  The new {cmd:cluster medianlinkage} command provides median linkage
    hierarchical clustering, also known as Gower's method.  See help
    {help clmedian}.

{p 6 10 2}
5.  The new {cmd:cluster stop} command provides stopping rules.  Two popular
    stopping rules are provided, the Calinski & Harabasz pseudo-F index
    (Calinski and Harabasz (1974)) and the Duda & Hart Je(2)/Je(1) index with
    associated pseudo T-squared (Duda and Hart (1973)).  See help
    {help clstop}.

{p 10 10 2}
    Additional stopping rules can be added; see help {help clprog}.

{p 6 10 2}
6.  Two new dissimilarity measures have been added:  {cmd:L2squared} and
    {cmd:Lpower(}{it:#}{cmd:)}.  {cmd:L2squared} provides squared Euclidean
    distance.  {cmd:Lpower(}{it:#}{cmd:)} provides the Minkowski distance
    metric with argument {it:#} raised to the {it:#} power.  See help
    {help cldis}.

{p 6 10 2}
7.  A list of the variables used in the cluster analysis is now saved with the
    cluster analysis structure, which is useful for programmers; see help
    {help clprog}.


{title:What's new in statistics useful in all fields}

{p 6 10 2}
1.  The following new estimators are available:

{p 10 14 2}
    a.  {cmd:manova} fits multivariate analysis-of-variance (MANOVA) and
	multivariate analysis-of-covariance (MANCOVA) models for balanced and
	unbalanced designs, including designs with missing cells; and for
	factorial, nested, or mixed designs.  See help {help manova}.
	({cmd:manovatest} provides multivariate tests involving terms from the
	most recently fitted {cmd: manova}; see help {help manovatest}.)

{p 10 14 2}
    b.  {cmd:rologit} fits the rank-order logit model, also known as the
	exploded logit model.  This model is a generalized McFadden's choice
	model as fitted by {cmd:clogit}.  In the choice model, only the
	alternative that maximizes utility is observed.  {cmd:rologit} fits
	the corresponding model in which the preference ranking of the
	alternatives is observed, not just the alternative that is ranked
	first.  {cmd:rologit} supports incomplete rankings and ties
	("indifference").  See help {help rologit}.

{p 10 14 2}
    c.  {cmd:frontier} fits stochastic frontier models with technical or cost
	inefficiency effects.  {cmd:frontier} can fit models in which the
	inefficiency error component is assumed to be from one of the three
	distributions: half-normal, exponential, or truncated-normal.  In
	addition, when the inefficiency term is assumed to be either
	half-normal or exponential, {cmd:frontier} can fit models in which the
	error components are heteroskedastic, conditional on a set of
	covariates.  {cmd:frontier} can also fit models in which the mean of
	the inefficiency term is modeled as a linear function of a set of
	covariates.  See help {help frontier}.

{p 10 10 2}
These new estimators are in addition to the new estimators listed in previous
sections.

{p 6 10 2}
2.  New command {cmd:mfp} selects the fractional polynomial model that best
    predicts the dependent variable from the independent variables; see help
    {help mfp}.

{p 6 10 2}
3.  The new {cmd:nlcom} command computes point estimates, standard errors, t
    and Z statistics, p-values, and confidence intervals for nonlinear
    combinations of coefficients after any estimation command.  Results are
    displayed in the table format that is commonly used for displaying
    estimation results.  The standard errors are based on the delta method, an
    approximation appropriate in large samples.  See help {help nlcom}.

{p 6 10 2}
4.  The new {cmd:predictnl} command produces nonlinear predictions after any
    Stata estimation command, and optionally, can calculate the variance,
    standard errors, Wald test-statistics, significance levels, and point-wise
    confidence intervals for these predictions.  Unlike {cmd:testnl} and
    {cmd:nlcom}, the quantities generated by {cmd:predictnl} are allowed to
    vary over the observations in the data.  The standard errors and other
    inference-related quantities are based on the "delta method", an
    approximation appropriate in large samples.  See help {help predictnl}.

{p 6 10 2}
5.  The new {cmd:bootstrap} command replaces the old {cmd:bstrap} and {cmd:bs}
    commands.  {cmd:bootstrap} has an improved syntax and allows for
    stratified sampling.  See help {help bootstrap}.

{p 10 10 2}
    Existing command {cmd:bsample} also now accepts the {cmd:strata()} option,
    and it has a new {cmd:weight()} option that allows the user to save the
    sample frequency instead of changing the data in memory.  See help
    {help bootstrap}.

{p 6 10 2}
6.  The existing {cmd:bstat} command can now construct bias-corrected and
    accelerated (BCa) confidence intervals.  In addition, {cmd:bstat} is now
    an e-class command, meaning all the post-estimation commands can be used
    on bootstrap results.  See help {help bootstrap}.

{p 6 10 2}