Re: st: maximum likelihood no observations error

[Nick Cox <njcoxstata@gmail.com>]

Generic advice on Statalist, which you should have read in the FAQ.
Please give full references, not minimal references such as Poi
(2002).

Specific advice on your problem. Even experienced programmers often do
not have much success in just looking at a program to find bugs. It's
best to be proactive. -list- the data at some point and then identify
what is not present that should be, or missing. If your real dataset
is big, it might be best to test the program on a small subset.

Nick

On Tue, Mar 12, 2013 at 8:41 AM, andylaustata <andylaustata@126.com> wrote:

> Learning from Poi (2002)'s QUAIDS ml evaluator, I come up with my own ml evaluator for a new demand system that I'm currently working on.
>
> But, while I do the ml check, it always gets stuck at Test 7 and reports no observations error (i.e. error 2000). I'm copying the code here. If anyone could give me some help, I would appreciate it very much.
>
> By the way, if you need some data to try this ml evaluator, the food data used in Poi (2002) might be a good choice, so basically, you just type "use food, clear".
>
> Cheers,
> Ou
>
> program define RDS1
>
>    version 11.1
>    args todo b lnf
>    //wi: shares; pi: prices; expfd: total expenditure
>    quietly{
>
>           tempname gammasum tausum etasum alpha_v beta_v theta_v gamma_v tau_v eta_v alpha beta theta gamma tau eta
>           tempvar P1 P2 P3 R1 R2 R3
>
>           local nm = $NEQN-1
>
>           scalar `alpha_v' = `b'[1,1]
>           scalar `beta_v' = `b'[1,2]
>           scalar `theta_v' = `b'[1,3]
>           matrix `gamma_v' = `b'[1,(3+1)..(3+`nm')]
>           matrix `tau_v' = `b'[1,(3+`nm'+1)..(3+2*`nm')]
>           matrix `eta_v' = `b'[1,(3+2*`nm'+1)..(3+3*`nm')]
>
>           scalar `alpha' = exp(`alpha_v')
>           scalar `beta' = invlogit(`beta_v')
>           scalar `theta' = invlogit(`theta_v')
>           matrix `gamma' = J(1,$NEQN,1)
>           matrix `tau' = J(1,$NEQN,1)
>           matrix `eta' = J(1,$NEQN,1)
>
>           scalar `gammasum' = 0 //the sum in the reparameterization of gamma
>           forvalues i = 1/`nm' {
>              scalar `gammasum' = `gammasum' + exp(`gamma_v'[1,`i'])
>           }
>
>           matrix `gamma'[1,1] = 1/(1+`gammasum')
>           forvalues i = 2/$NEQN {
>              matrix `gamma'[1,`i'] = exp(`gamma_v'[1,(`i'-1)])/(1+`gammasum')
>           }
>
>           scalar `tausum' = 0 //the sum in the reparameterization of tau
>           forvalues i = 1/`nm' {
>              scalar `tausum' = `tausum' + exp(`tau_v'[1,`i'])
>           }
>
>           matrix `tau'[1,1] = 1/(1+`tausum')
>           forvalues i = 2/$NEQN {
>              matrix `tau'[1,`i'] = exp(`tau_v'[1,(`i'-1)])/(1+`tausum')
>           }
>
>
>           scalar `etasum' = 0 //the sum in the reparameterization of eta
>           forvalues i = 1/`nm' {
>              scalar `etasum' = `etasum' + exp(`eta_v'[1,`i'])
>           }
>
>           matrix `eta'[1,1] = 1/(1+`etasum')
>           forvalues i = 2/$NEQN {
>              matrix `eta'[1,`i'] = exp(`eta_v'[1,(`i'-1)])/(1+`etasum')
>           }
>
>           /* Form the price index variable P1.*/
>           gen double `P1' = p1^`gamma'[1,1]
>           forvalues i = 2/$NEQN{
>              replace `P1' = `P1'*(p`i'^`gamma'[1,`i'])
>           }
>
>           /* Form the price index variable P2.*/
>           gen double `P2' = p1^`tau'[1,1]
>           forvalues i = 2/$NEQN{
>              replace `P2' = `P2'*(p`i'^`tau'[1,`i'])
>           }
>
>           /* Form the price index variable P3.*/
>           gen double `P3' = p1*`eta'[1,1]
>           forvalues i = 2/$NEQN{
>              replace `P3' = `P3'+(p`i'*`eta'[1,`i'])
>           }
>
>           /*gen R1*/
>           gen double `R1' = `alpha'*expfd^`alpha'*`P1'^(-`alpha'-1)
>
>           /*gen R2*/
>           gen double `R2' = `beta'*expfd^(-`beta')*`P2'^(`beta'-1)
>
>           /*gen R3*/
>           gen double `R3' = `theta'*expfd^(-`theta')*`P3'^(`theta'-1)
>
>           /*Now generate the error terms.*/
>           forvalues i = 1/`nm' {
>              tempvar lnl_eps`i'
>                  gen double `lnl_eps`i'' = w`i' - (`gamma'[1,`i']*`R1'*`P1'+`tau'[1,`i']*`R2'*`P2'+`eta'[1,`i']*`R3'*p`i')/(`R1'*`P1'+`R2'*`P2'+`R3'*`P3')
>           }
>
>           local allofem ""
>           forvalues i = 1/`nm' {
>              local allofem "`allofem' `lnl_eps`i''"
>           }
>
>           /*Form sigma.*/
>           matrix accum sigma = `allofem', noconstant
>           local nobs = r(N)
>           matrix sigma = sigma/`nobs'
>
>            /* Finally, compute the likelihood function. */
>       scalar `lnf' = -1*`nobs'/2*(`nm'*(1+ln(2*_pi)) + ln(det(sigma)))
>    }
>
> end
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/