nlcom.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

{smcl}
{* 14mar2005}{...}
{cmd:help nlcom} {right:dialog:  {bf:{dialog nlcom}}}
{hline}

{title:Title}

{p2colset 5 18 20 2}{...}
{p2col :{hi:[R] nlcom} {hline 2}}Nonlinear combinations of estimators{p_end}
{p2colreset}{...}


{title:Syntax}

{phang}
Nonlinear combination of estimators{hline 2}one expression

{p 8 15 2}
{cmd:nlcom} [{it:name}{cmd::}]{it:{help exp}} [{cmd:,} {it:options}]

{phang}
Nonlinear combinations of estimators{hline 2}more than one expression

{p 8 15 2}
{cmd:nlcom} {cmd:(}[{it:name}{cmd::}]{it:{help exp}}{cmd:)}
[{cmd:(}[{it:name}{cmd::}]{it:exp}{cmd:)} ...] 
[{cmd:,} {it:options}]

{phang}
The second syntax means that if more than one expression is specified, each
must be surrounded by parentheses.  {it:exp} is any function of the parameter
estimates that is valid syntax for {helpb testnl}. Note, however, that
{it:exp} may not contain an equal sign or a comma.  The optional {it:name} is
any valid Stata name and labels the transformation.

{synoptset 12}{...}
{synopthdr}
{synoptline}
{synopt :{opt post}}post estimation results{p_end}
{synopt :{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}
{synopt :{opt iter:ate(#)}}maximum number of iterations{p_end}
{synoptline}
{p2colreset}{...}


{title:Description}

{pstd}
{cmd:nlcom} computes point estimates, standard errors, test statistics,
significance levels, and confidence intervals for (possibly) nonlinear combinations of
parameter estimates after any Stata estimation command.  Results are displayed
in the usual table format used for displaying estimation results.
Calculations are based on the "delta method", an approximation appropriate in
large samples.

{pstd}
{cmd:nlcom} supports {helpb svy} estimation commands ({opt "svy: regress"},
{opt "svy: logit"}, etc.).


{title:Options}

{phang}
{opt post} causes {cmd:nlcom} to behave like a Stata estimation ({cmd:eclass})
command.  When {opt post} is specified, {cmd:nlcom} will post the vector of
transformed estimators and its estimated variance-covariance matrix to
{cmd:e()}. This option, in essence, makes the transformation permanent.  Thus
you could, after {opt post}ing, treat the transformed estimation results in
the same way as you would treat results from other Stata estimation commands.
For example, after posting, you could redisplay the results by typing
{cmd:nlcom} without any arguments, or use {helpb test} to perform simultaneous
tests of hypotheses on linear combinations of the transformed estimators.

{pmore}
Note that specifying {opt post} clears out the previous estimation results,
which can be recovered only by refitting the original model or by storing the
estimation results before running {cmd:nlcom} and then restoring them; see
{helpb estimates}.

{phang}
{opt level(#)} specifies the confidence level, as a percentage,
for confidence intervals.  The default is {cmd:level(95)} or as set by
{helpb set level}.

{phang}
{opt iterate(#)} specifies the maximum number of
iterations used to find the optimal step size in calculating
numerical derivatives of the transformations with respect to the original
parameters.  By default, the maximum number of iterations is 100,
but convergence is usually achieved only after a few iterations.  You should
rarely have to use this option.


{title:Comparison with lincom}

{pstd}
{cmd:nlcom} is a generalization of {helpb lincom} that allows the estimation of
nonlinear transformations of model parameters.  In cases where you are
estimating a single transformation and that transformation is linear, use
{helpb lincom}; it is faster.  However, when estimating more than one linear
transformation or combinations of linear and nonlinear transformations, using
{cmd:nlcom} has the added benefit that you can obtain the variance-covariance
matrix (which is saved in {cmd:r(V)}) of the joint transformation.  Note that
{helpb lincom} does not allow the simultaneous estimation of multiple linear
combinations.


{title:Remark on the manipulability of nonlinear Wald tests}

{pstd}
In contrast to likelihood-ratio tests, different{hline 2}mathematically
equivalent{hline 2}formulations of a hypothesis may lead to different results
for a nonlinear Wald test (lack of "invariance"). For instance, the two
hypotheses

	H0: {it:coefficient} = 0

	H0: exp({it:coefficient}) - 1 = 0

{pstd}
are mathematically equivalent expressions but do not yield the same test
statistic and p-value. In extreme cases, under one formulation, one would
reject H0 while under an equivalent formulation one would not reject H0.


{title:Examples}

{phang}{cmd:. nbreg y}{p_end}
{phang}{cmd:. nlcom exp(_b[_cons])}{p_end}
{phang}{cmd:. nlcom mean:exp(_b[_cons])}{p_end}

{phang}{cmd:. regress y x1 x2 x3 z1 z2}{p_end}
{phang}{cmd:. nlcom log(3*_b[x1] + 500*_b[x3])}{p_end}
{phang}{cmd:. nlcom (t1:_b[x1] - _b[x2]) (t2:_b[z1] - _b[z2])}{p_end}
{phang}{cmd:. mat list r(V)}{p_end}
{phang}{cmd:. nlcom (t1:_b[x1] - _b[x2]) (t2:_b[z1] - _b[z2]), post}{p_end}
{phang}{cmd:. correlate, _coef}{p_end}

{phang}{cmd:. mlogit insure age male nonwhite site2 site3}{p_end}
{phang}{cmd:. nlcom [Prepaid]_b[male] / [Prepaid]_b[nonwhite] - 1}


{title:Also see}

{psee}
Manual:  {bf:[R] nlcom}

{psee}
Online:  {helpb lincom}, {helpb predictnl}, {helpb test}, {helpb testnl}
{p_end}