Re: st: AW: Predict command

[Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>]

I fully agree with Nick and Steve - you should probably not predict
from a model by posting restrictions post-hoc. However, it is possible
that interesting statistical properties (inconsistency results) can be
derived. There are many ways to do what you want, and this in an
ungainly way using Mata:
********************************************
clear*
clear mata
sysuse auto
regress mpg weight c.weight#c.weight foreign turn

mata
vPval = 2*ttail(st_numscalar("e(df_r)"),abs(st_matrix("e(b)")'
	:/ diagonal(cholesky(diag(st_matrix("e(V)"))))))
vSpval = mm_cond(vPval :< .05, 1, 0) // need Ben Jann's -moremata- package
vBetaSelect = vSpval :* st_matrix("e(b)")'
st_view(mData=., ., (st_macroexpand(regexr(regexr("`e(cmdline)'",
"`e(depvar)'", ""), "`e(cmd)'", ""))))
	// this will choke on options
mData = mData, J(rows(mData),1,1)
vLinPredSelect = mData*vBetaSelect
// compare with the full linear predictor
stata("predict vLinPred, xb")
st_view(vLinPred = ., ., "vLinPred")
mm_plot((vLinPredSelect, vLinPred)) // need Ben Jann's -moremata- package
end
getmata vLinPredSelect
********************************************

T

PS. The p-value line is due to M. Buis's article (Stata tip 53: Where
did my p-values go?) in SJ (Vol. 7, No. 4).



2010/4/21 Steve Samuels <sjsamuels@gmail.com>:
> I'm not Nick, but: No.
>
> On Wed, Apr 21, 2010 at 9:14 AM, Rasika Raghavan <rasikarag@gmail.com> wrote:
>> Nick,
>>
>> Anyway I can do so without re-running the model?
>>
>> Rasika
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/statalist/faq
>> *   http://www.ats.ucla.edu/stat/stata/
>>
>
>
>
> --
> Steven Samuels
> sjsamuels@gmail.com
> 18 Cantine's Island
> Saugerties NY 12477
> USA
> Voice: 845-246-0774
> Fax:    206-202-4783
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>



-- 
To every ω-consistent recursive class κ of formulae there correspond
recursive class signs r, such that neither v Gen r nor Neg(v Gen r)
belongs to Flg(κ) (where v is the free variable of r).

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/