RE: st: NLSUR Quaids

[Nick Cox <n.j.cox@durham.ac.uk>]

This line is wrong 

scalar `a5' = 1 - `a1' - `a2' - `a3' - `a0'

`a0' is not defined. I guess you mean `a4'. 

Nick 
n.j.cox@durham.ac.uk 


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Rahkovsky, Ilya
Sent: 27 October 2011 18:52
To: statalist@hsphsun2.harvard.edu
Subject: RE: st: NLSUR Quaids

Brian thank you for very much for your response. I have tried it and now I am getting an error 199. I looked at the previous discussion about this error, but it seems that I don't have the issues discussed.  Please let me know if there are any other errors there. 


CODE 

*
set trace on
set tracedepth 4
*/



use "E:\Data\Demand\agg_aids.dta", clear

gen exp = s0 + s1 + s2 + s3 + s9
gen lnexp =ln(exp)

gen lnp1 = ln(p0)
gen lnp2 = ln(p1)
gen lnp3 = ln(p2)
gen lnp4 = ln(p3)
gen lnp5 = ln(p9)

drop w*

gen w1 = s0 / exp
gen w2 = s1 / exp
gen w3 = s2 / exp
gen w4 = s3 / exp
gen w5 = s9 / exp

 

keep w1 w2 w3 w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnexp 
*/

*program drop nlsuraids
program nlsuraids
        version 10.1
syntax varlist(min=10 max=10) if, at(name) 

tokenize `varlist'

args w1 w2 w3 w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnexp 

  
tempname a1 a2 a3 a4 a5

scalar `a1' = `at'[1,1] 

scalar `a2' = `at'[1,2] 

scalar `a3' = `at'[1,3] 

scalar `a4' = `at'[1,4] 

scalar `a5' = 1 - `a1' - `a2' - `a3' - `a0'

  

tempname b1 b2 b3 b4 b5 

scalar `b1' = `at'[1,5] 

scalar `b2' = `at'[1,6] 

scalar `b3' = `at'[1,7] 

scalar `b4' = `at'[1,8] 
  
scalar `b5' = -`b1' - `b2' - `b3' - `b4'

  

tempname g11 g12 g13 g14 g15

tempname g21 g22 g23 g24 g25

tempname g31 g32 g33 g34 g35

tempname g41 g42 g43 g44 g44

tempname g51 g52 g53 g54 g55

  
scalar `g11' = `at'[1,9] 

scalar `g12' = `at'[1,10] 

scalar `g13' = `at'[1,11] 

scalar `g14' = `at'[1,12]  

scalar `g15' = -`g11' - `g12' - `g13' - `g14'

  

scalar `g21' = `g12' 

scalar `g22' = `at'[1,13] 

scalar `g23' = `at'[1,14] 

scalar `g24' = `at'[1,15]

scalar `g25' = -`g21' - `g22' - `g23' - `g24'

  
scalar `g31' = `g13' 

scalar `g32' = `g23' 

scalar `g33' = `at'[1,16] 

scalar `g34' = `at'[1,17]

scalar `g35' = -`g31' - `g32' - `g33' - `g34'
 

scalar `g41' = `g14' 

scalar `g42' = `g24' 

scalar `g43' = `g34' 

scalar `g44' = `at'[1,18]

scalar `g45' = -`g41' - `g42' - `g43' - `g44'  

  
scalar `g51' = `g15' 

scalar `g52' = `g25' 

scalar `g53' = `g35' 

scalar `g54' = `g45'

scalar `g55' = -`g51' - `g52' - `g53' - `g54' 
  

quietly { 

tempvar lnpindex 

gen double `lnpindex' = 5 + `a1'*`lnp1' + `a2'*`lnp2' + `a3'*`lnp3' + `a4'*`lnp4' + `a5'*`lnp5'

forvalues i = 1/5 { 

forvalues j = 1/5 { 

replace `lnpindex' = `lnpindex' + 0.5*`g`i'`j''*`lnp`i''*`lnp`j'' 

} 

} 

replace `w1' = `a1' + `g11'*`lnp1' + `g12'*`lnp2' + `g13'*`lnp3' + `g14'*`lnp4' + `g15'*`lnp5' + `b1'*(`lnexp' - `lnpindex') 

replace `w2' = `a2' + `g21'*`lnp1' + `g22'*`lnp2' + `g23'*`lnp3' + `g24'*`lnp4' + `g25'*`lnp5' + `b2'*(`lnexp' - `lnpindex') 

replace `w3' = `a3' + `g31'*`lnp1' + `g32'*`lnp2' + `g33'*`lnp3' + `g34'*`lnp4' + `g35'*`lnp5' + `b3'*(`lnexp' - `lnpindex') 

replace `w4' = `a4' + `g41'*`lnp1' + `g42'*`lnp2' + `g43'*`lnp3' + `g44'*`lnp4' + `g45'*`lnp5' + `b4'*(`lnexp' - `lnpindex') 
} 
 

end 

nlsur aids @ w1 w2 w3 w4  lnp1 lnp2 lnp3 lnp4 lnp5 lnexp, ifgnls nequations(4) param(a1 a2 a3 a4 b1 b2 b3 b4  g11 g12 g13 g14 g22 g23 g24 g33 g34 g44 ) nolog


ERROR


      ------------------------------------------------------------------ begin nlsuraids ---
      - version 10.1
      - syntax varlist(min=10 max=10) if, at(name)
      - tokenize `varlist'
      = tokenize __00000D __00000E __00000F __00000G lnp1 lnp2 lnp3 lnp4 lnp5 lnexp
      - args w1 w2 w3 w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnexp
      - tempname a1 a2 a3 a4 a5
      - scalar `a1' = `at'[1,1]
      = scalar __00000H = __000003[1,1]
      - scalar `a2' = `at'[1,2]
      = scalar __00000I = __000003[1,2]
      - scalar `a3' = `at'[1,3]
      = scalar __00000J = __000003[1,3]
      - scalar `a4' = `at'[1,4]
      = scalar __00000K = __000003[1,4]
      - scalar `a5' = 1 - `a1' - `a2' - `a3' - `a0'
      = scalar __00000L = 1 - __00000H - __00000I - __00000J - 
      -------------------------------------------------------------------- end nlsuraids ---
    - }
    - if _rc {
    - di as error "nlsur`eqn' returned " _rc
    = di as error "nlsuraids returned " _rc
nlsuraids returned 198
    - di as error "verify that nlsur`eqn' is a function evaluator program"
    = di as error "verify that nlsuraids is a function evaluator program"
verify that nlsuraids is a function evaluator program
    - exit _rc
      }
      }
    ----------------------------------------------------------------- end nlsur.Estimate ---
  ---------------------------------------------------------------------------- end nlsur ---
r(198);

end of do-file

r(198);


Ilya


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Brian P. Poi
Sent: Thursday, October 27, 2011 11:36 AM
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: NLSUR Quaids

Rahkovsky, Ilya wrote:
> Hello,
> 
> I am trying to follow Brians Poi (2008) QAIDS model. I keep getting an error 111. I have a trace on, and I guess that the error is in the specifying the number of equations, but I am not sure what I should change in the code.
> 
> Thank you, Ilya
...
> MY CODE 
> 
> set trace on
> set tracedepth 4
> 
> program drop nlsuraids
> program nlsuraids
>         version 10.1
> use "E:\Data\Demand\agg_aids.dta", clear
> 
> 
> gen exp = s0 + s1 + s2 + s3 + s9
> gen lnexp =ln(exp)
> 
> gen lnp1 = ln(p0)
> gen lnp2 = ln(p1)
> gen lnp3 = ln(p2)
> gen lnp4 = ln(p3)
> gen lnp5 = ln(p9)
> 
> drop w*
> 
> gen w1 = s0 / exp
> gen w2 = s1 / exp
> gen w3 = s2 / exp
> gen w4 = s3 / exp
> gen w5 = s9 / exp
> 
> keep w1 w2 w3 w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnexp 
> 
> syntax varlist(min=10 max=10) if, at(name) 
...


The problem is that the way your program is written, you are reloading the dataset every time the program nlsuraids gets called.  Instead, you should load in the data once, define the nlsuraids program to just compute the predicted share equations, then call -nlsur- The solution is simple.  In your sample code, just move the three lines

program drop nlsuraids
program nlsuraids
         version 10.1

down, just before the -syntax- statement.

The other problem is that your -tokenize- statement is not correct.

The outline of your entire do-file should be something like


use "E:\Data\Demand\agg_aids.dta", clear
gen exp = ...
gen lnexp = ln(exp)
gen lnp1 = ...
...
gen w1 = ...


capture program drop nlsuraids
program nlsuraids
         version 10.1
         syntax varlist(min=10 max=10) if, at(name)

         tokenize `varlist'
	args w1 w2 w3 w4 lnp1 lnp2 lnp3 lnp4 lnp5 lnexp

         ...
         <I haven't studied the rest of your code carefully, but it looks okay>
         ...
         replace `w1' ...
         replace `w2' ...
         replace `w3' ...
         replace `w4' ...

end

nlsur aids @ w1 w2 w3 w4  lnp1-lnp5 lnexp, ifgnls nequations(4) param(a1 a2 a3 a4 b1 b2 b3 b4  g11 g12 g13 g14 g22 g23 g24 g33 g34 g44 ) nolog

*
*   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/