stpm2 - r(111) - predict hr2, hrnumerator --> if not found

[Denise Vella]

I am just doing a few exercises using P Dickman's steps - when predicting hr using hrnumerator
Stata gives an error : r(111) if not found
Why is this happening?

clear all
use https://pauldickman.com/data/melanoma.dta if stage>0, clear

// rename to a more descriptive name
rename year8594 period

// create dummy variable for sex
generate male=(sex==1)
label define male 0 "Female" 1 "Male"
label values male male

// stset for cause-specific survival (status==1 is death due to melanoma) 
//survm - survival time, status 12 - death 
stset surv_mm, fail(status==1) scale(12) exit(time 120)

range temptime 0 10 51 //creating a time variable temptime 
predict hr, hrnumerator(male 1) ci timevar(temptime) 

//Trying to create like empirical but instead for females
predict hr2, hrnumerator(male 0) ci timevar(temptime) 
predict hr2, hrnumerator(male 1) ci timevar(temptime)     ///trying to create HR for females

Stata gives the following error, why is this?

predict hr2, hrnumerator(male 0) ci timevar(temptime)
if not found
r(111);

end of do-file


I will post another post on the method on P dICKMAN website

[Bjarte Aagnes]

Revise your example code, it does not produce the reported error. (the error code suggest some misstyping of a variable name)

[Denise Vella]

Thanks for explaining what it means
but I haven't mistyped any variable unless of course you can see something abnormal?

clear all
use https://pauldickman.com/data/melanoma.dta if stage>0, clear

// rename to a more descriptive name
rename year8594 period

// create dummy variable for sex
generate male=(sex==1)
label define male 0 "Female" 1 "Male"
label values male male

// stset for cause-specific survival (status==1 is death due to melanoma) 
//survm - survival time, status 12 - death 
stset surv_mm, fail(status==1) scale(12) exit(time 120)

range temptime 0 10 51 //creating a time variable temptime 

// estimate the time-varying hazard ratio using flexible parametric models
stpm2 male i.period i.agegrp i.stage, scale(h) df(5) eform tvc(stage) dftvc(3)


//Trying to create like empirical but instead for females
predict hr2, hrnumerator(male 0) ci timevar(temptime) 
predict hr2, hrnumerator(male 1) ci timevar(temptime)     ///trying to create HR for females