how to loop two parameters for self-build maximum likelihood program?

[Ting Meng]

Dear all,

I'd like loop both d and theta (i.e., d=0.1(0.05)0.2, and theta=1(0.1)2) for the following program myest. I tried "forvalues after program" and "program within forvalues". Neither works.
Would you please help to figure it out? Thank you.


*****code as below
capture program drop myest
program define myest
version 14
args lnf Xb
tempvar d theta
quietly generate double `d'=0.1
quietly generate double `theta'=1
quietly replace `lnf' = ln(normal((ln(300/((1-exp(-1*`d'-0.1))/(`d'+0.1)*exp(-1*`d'*2)))-`Xb')/`theta')) if $ML_y1==2
quietly replace `lnf' = ln(1-normal((ln(300/((1-exp(-1*`d'-0.1))/(`d'+0.1)*exp(-1*`d'*19)))-`Xb')/`theta')) if $ML_y1==20
quietly replace `lnf' = ln(normal((ln(300/exp(-1*$ML_y1*`d')*((1-exp(-1*`d'-0.1))/(`d'+0.1)))-`Xb')/`theta')-normal((ln(300/exp(-1*($ML_y1-1)*`d')*((1-exp(-1*`d'-0.1))/(`d'+0.1)))-`Xb')/`theta')) if ($ML_y1<20)&($ML_y1>2)
end

qui ml model lf myest (y= x1 x2)
qui ml check
qui ml maximize
display e(ll)

[George Ford]

forv d = 0.1(0.05)0.2 {
     forv theta = 1(0.1)2)
     *do stuff with `d' and `theta'
}
}

[FernandoRios]

Perhaps it’s better to ask…what is the goal?
are d and theta added parameters for maximization?
or do you want to see sensitive to them
that may help giving a good answer

[Ting Meng]

forv d = 0.1(0.05)0.2 {
forv theta = 1(0.1)2)
*do stuff with `d' and `theta'
}
}

Thank you George for your prompt response. I tried similar coding as below, but the STATA report r(100) "varlist required".

capture program drop myest
program define myest
version 14
args lnf Xb
tempvar dd theta
quietly replace `lnf' = ln(normal((ln(300/((1-exp(-1*`dd'-0.1))/(`dd'+0.1)*exp(-1*`dd'*2)))-`Xb')/`theta')) if $ML_y1==2
quietly replace `lnf' = ln(1-normal((ln(300/((1-exp(-1*`dd'-0.1))/(`dd'+0.1)*exp(-1*`dd'*19)))-`Xb')/`theta')) if $ML_y1==20
quietly replace `lnf' = ln(normal((ln(300/exp(-1*$ML_y1*`dd')*((1-exp(-1*`dd'-0.1))/(`dd'+0.1)))-`Xb')/`theta')-normal((ln(300/exp(-1*($ML_y1-1)*`dd')*((1-exp(-1*`dd'-0.1))/(`dd'+0.1)))-`Xb')/`theta')) if ($ML_y1<20)&($ML_y1>2)
end

forvalues dd = 0.1(0.05)0.2 {
forvalues theta = 1(0.1)2 {
myest
ml model lf myest (y1= x1 x2)
ml maximize
display `dd'
display `theta'
display e(ll)
}
}

[Ting Meng]

Perhaps it’s better to ask…what is the goal?
are d and theta added parameters for maximization?
or do you want to see sensitive to them
that may help giving a good answer

Thank you Fernando. You made a great point.
The goal is to estimate d, theta, and b(the coeffcients) together from maximizing the log-likelihood function. However, by Program myest, I only can estimate b(the coeffcients) given the value of d and theta. Therefore, as a second best way, I'd explore the space of d and theta (d=0.1(0.05)0.2, theta=1(0.1)2), and check which combination of d and theta has the highest log-likelihood value.

[FernandoRios]

I see
In that case, my suggestion is
1) modify your code so all parameters (d theta and b) are estimated in the model.
2) use constraints to change the values for d and theta

F

[George Ford]

I can't figure out what you're up to, but there are some issues I see.

First, the tempvar line is resetting `dd' and `theta' as tempvars (so you lose the loop values).

Second, you need to send `dd' and `theta' as arguments.

capture program drop myest
program define myest
version 14
args dd theta
di `dd' _col(20) `theta'
end

forvalues dd = 0.1(0.05)0.2 {
     forvalues theta = 1(0.1)2 {
          myest `dd' `theta'
}
}

Third, I'm not sure what you're doing with myest in the -ml model- line; you're calling a program. Where does `Inf' fit into this?