asmprobit.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}
{* *! version 1.0.0  16may2005}{...}
{cmd:help asmprobit} {right:dialog:  {bf:{dialog asmprobit}}{space 15}}
{right:also see:  {help asmprobit postestimation}}
{hline}

{title:Title}

{p2colset 5 22 24 2}{...}
{synopt :{hi:[R] asmprobit} {hline 2}}Maximum simulated-likelihood alternative-specific multinomial probit models
{p_end}
{p2colreset}{...}


{title:Syntax}

{p 8 18 2}
{cmd:asmprobit} 
{depvar} 
[{indepvars}] 
{ifin}
{weight}{cmd:,}
{opt case(varname)}
{opt alt:ernatives(varname)}
[{it:options}]

{synoptset 28 tabbed}{...}
{synopthdr}
{synoptline}
{syntab:Model}
{p2coldent :* {opth case(varname)}}use {it:varname} to identify cases{p_end}
{p2coldent :* {opth alt:ernatives(varname)}}use {it:varname} to identify the alternatives 
available for each case{p_end}
{synopt : {opth casev:ars(varlist)}}case-specific variables{p_end}
{synopt :{cmdab:const:raints(}{it:{help estimation options##constraints():constraints}}{cmd:)}}apply specified linear constraints{p_end}

{syntab :Model 2}
{synopt :{opth corr:elation(asmprobit##cortype:correlation)}}correlation structure of the latent-variable errors{p_end}
{synopt :{opth std:dev(asmprobit##stdtype:stddev)}}variance structure of the latent-variable errors{p_end}
{synopt :{opt nocon:stant}}suppress the constant terms{p_end}
{synopt :{opt base:alternative(#|lbl|str)}}alternative used for normalizing location{p_end}
{synopt :{opt scale:alternative(#|lbl|str)}}alternative used for normalizing scale{p_end}

{syntab :SE/Robust}
{synopt :{opth vce(vcetype)}}type of VCE; {it:vcetype} may be {opt oim},
{opt r:obust}, {opt opg}, {opt boot:strap}, or {opt jack:knife}{p_end}
{synopt :{opt r:obust}}synonym for {cmd:vce(robust)}{p_end}
{synopt :{opth cl:uster(varname)}}adjust standard errors for intragroup correlation{p_end}

{syntab :Reporting}
{synopt :{opt notran:sform}}prevents reparameterizing the Cholesky factored variance-covariance
estimates to the standard deviation and correlation metric{p_end}
{synopt :{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}

{syntab :Int options}
{synopt :{opth intm:ethod(asmprobit##seqtype:seqtype)}}type of quasi- or pseudo-uniform sequence{p_end}
{synopt :{opt intp:oints(#)}}number of points in the sequence{p_end}
{synopt :{opt intb:urn(#)}}starting index in the Hammersley or Halton sequence{p_end}
{synopt :{opth ints:eed(generate##seed:code)}}uniform random number generator seed{p_end}
{synopt :{opt anti:thetics}}use antithetic draws{p_end}
{synopt :{opt initb:hhh(#)}}use the bhhh optimization algorithm for the first {it:#} 
iterations{p_end}

{syntab :Max options}
{synopt :{it:{help maximize}}}control the maximization process{p_end}
{synoptline}
{p2colreset}{...}

{synoptset 23}{...}
{marker cortype}
{synopthdr :correlation}
{synoptline}
{synopt :{opt uns:tructured}}one correlation parameter for each pair of 
alternatives; correlations with the {opt basealternative()} are zero{p_end}
{synopt :{opt exc:hangeable}}one correlation parameter common to all pairs of 
alternatives; correlations with the {opt basealternative()} are zero{p_end}
{synopt :{opt ind:ependent}}constrain all correlation parameters to zero{p_end}
{synopt :{opt pat:tern} {it:matname}}user-specified matrix identifying the
correlation pattern{p_end}
{synopt :{opt fix:ed} {it:matname}}user-specified matrix identifying the free
and fixed correlation parameters{p_end}
{synoptline}
{p2colreset}{...}

{synoptset 23}{...}
{marker stdtype}
{synopthdr :stddev}
{synoptline}
{synopt :{opt het:eroskedastic}}standard deviation estimate for each alternative;  
standard deviations for {opt basealternative()} and {opt scalealternative()} 
are one{p_end}
{synopt :{opt hom:oskedastic}}all standard deviations are one{p_end}
{synopt :{opt pat:tern} {it:matname}}user-specified matrix identifying 
the standard deviation pattern{p_end}
{synopt :{opt fix:ed} {it:matname}}user-specified matrix identifying 
the fixed and free standard deviations{p_end}
{synoptline}
{p2colreset}{...}
 
{synoptset 23}{...}
{marker seqtype}
{synopthdr :seqtype}
{synoptline}
{synopt :{opt ham:mersley}}Hammersley sequence{p_end}
{synopt :{opt hal:ton}}Halton sequence{p_end}
{synopt :{opt ran:dom}}uniform pseudorandom sequence{p_end}
{synoptline}
{p2colreset}{...}
{p 4 6 2}* {opt cases(varname)} and {opt alternatives(varname)} are required.{p_end}
{p 4 6 2}{cmd:bootstrap}, {cmd:by}, {cmd:jackknife}, {cmd:statsby}, and {cmd:xi} are allowed; see {help prefix}.{p_end}
{p 4 6 2}{cmd:fweight}s, {cmd:iweight}s, and {cmd:pweight}s are allowed; see {help weight}.{p_end}
{p 4 6 2}See {help asmprobit postestimation} for features available after estimation.{p_end}


{title:Description}

{pstd}
{cmd:asmprobit} fits multinomial probit models (MNP) using maximum simulated likelihood
(MSL).  The model allows you to relax the independence of irrelevant alternatives (IIA) 
property that is characteristic of the multinomial logistic model.  This is done by estimating the 
variance-covariance parameters of the latent-variable errors.  Each unique identifier
in the variable {opt case()} has multiple alternatives identified in the variable
{opt alternatives()}, and {it:depvar} indicates the selection made by each case.
{it:depvar} is coded zero or one, one indicating the chosen alternative and zero
indicating the alternatives not chosen. {cmd:asmprobit} allows for two types of 
independent variables: alternative specific variables, in which the values of 
each variable vary with each alternative; and case specific variables that vary 
with each case.


{title:Options}

{dlgtab:Model}

{phang}{opth case(varname)} specifies the variable that identifies each case.
{opt case()} is required.

{phang}{opth alternatives(varname)} specifies the variable that identifies the 
alternatives for each case.  The number of alternatives can vary with each case.  
{opt alternatives()} is required.

{phang}{opth casevars(varlist)} case specific variables that are constant
for each {opt case()}. If there are a maximum of J alternatives, then 
there will be J-1 sets of coefficients associated with {opt casevars()}.

{phang}
{opt constraints(constraints)}; see {help estimation options}.

{dlgtab:Model 2}

{phang}{opt correlation(correlation)}
specifies the correlation structure of the latent-variable errors.  
{opt correlation(unstructured)}, the default, is the most general and will result in
Jx(J-3)/2+1 correlation estimates.  The {opt correlation(exchangeable)}
provides for one correlation coefficient common to all latent-variables, with
the exception of the latent-variable associated with the {opt basealternative()}.
{opt correlation(independent)} assumes all correlations are zero.

{pmore}{cmd:correlation(pattern} {it:matname}{cmd:)} and 
{cmd:stddev(fixed} {it:matname}{cmd:)} 
gives you the flexibility in defining the correlation structure.  The matrix
{it:matname} is a J x J matrix.  For {cmd:correlation(pattern} {it:matname}{cmd:)}, 
{it:matname} contains sequential positive integers,
starting from 1, used to identify each correlation parameter: If there are 3
correlation parameters then they are identified by 1, 2, and 3.  The integers
can be repeated and missing values (.), or 0, indicate no correlation. For 
example, assume J = 4 with alternative levels 1, 2, 3, 4 and 
{cmd:basealternative(}{it:1}{cmd:)}.  Below are the unstructured and exchangeable 
correlation structures identified in the 4 by 4 lower triangular matrices.

{pmore}{space 9}unstructured{space 15}exchangeable{break}
{space 8}{c TLC}{space 12}{c TRC}{space 13}{c TLC}{space 12}{c TRC}{break} 
{space 5}{space 1}1{space 1}{c | } {bf:.}{space 2}{space 8}{c |}{space 10}{space 1}1{space 1}{c | } {bf:.}{space 2}{space 8}{c |}{break}
{space 5}{space 1}2{space 1}{c | } {bf:.}{space 2}{bf:.}{space 7}{c |}{space 10}{space 1}2{space 1}{c | } {bf:.}{space 2}{bf:.}{space 7}{c |}{break}
{space 5}{space 1}3{space 1}{c | } {bf:.}{space 2}1{space 2}{bf:.}{space 4}{c |}{space 10}{space 1}3{space 1}{c | } {bf:.}{space 2}1{space 2}{bf:.}{space 4}{c |}{break}
{space 5}{space 1}4{space 1}{c | } {bf:.}{space 2}2{space 2}3{space 2}{bf:.}{space 1}{c |}{space 10}{space 1}4{space 1}{c | } {bf:.}{space 2}1{space 2}1{space 2}{bf:.}{space 1}{c |}{break}
{space 8}{c BLC}{space 12}{c BRC}{space 13}{c BLC}{space 12}{c BRC}{break}