[Error Code] r(2000) (GLLAMM)

#1
I want to analyze using Multilevel Binary Logistic Regression modeling (with gllamm). I have 6.998 observation in level 1 and 15 province in level 2 (no missing data).

These are my variables:

y = smoking status (yes/no)

x1 = age (years)
x2 = duration of formal education years
x3 = media score (ratio)
x4 = working status (yes/no)
x5 = wealth index (1/2/3/4/5)
x6 = sex (male/female)
x7 = geography (urban/rural)
x12 = age (years) level-2
x22 = education (years) level-2
x32 = media (score) level-2
prov = province area

first i want to analyze null model with:
Code:
gllamm y, i(prov) link(logit) fam(binom) adapt nolog
but the result is
Code:
r(2000);
i already read the help for this error code, but it didn't give me any clue for what has been wrong.

then i checked with:
Code:
gllamm y, i(prov)
to test maybe i have some error from the data coding or related.

Those command gave the result (but i can't use the result, because i need link logit and binomial scale)

What specifically r(2000) error code mean? and how to deal with it?
is there any wrong with my data?

(I really confuse with this error, because i just want to check the NULL model not the complex one)

note: Stata v11 & v12 give similar result

**
my environment:
STATA SE v11 & v12, 4gb DDR2 PC6400, Intel Core2Duo 1,6ghz
Win7 Ultimate SP1 32bit
 
Last edited:
#3
yes, there is string variable in data file, variable = case_id.
But these variable didn't include in analysis.

is this problem from case_id variable?
 

bukharin

RoboStataRaptor
#5
Is y stored as a numeric variable only taking the values 0 or 1? If not, maybe -gllamm- is dropping those variables that don't contain 0/1 in y.
 
#6
Is y stored as a numeric variable only taking the values 0 or 1? If not, maybe -gllamm- is dropping those variables that don't contain 0/1 in y.
my "y" variable stored as numeric variable. what i mean before, as string variable "case_id" i already delete it but the result is the same.
 
#8
Yes, but is it a 0 or 1 variable? What happens when you type -codebook y-?
sorry for late response. Problem solved several days ago. My "y" variable coded with: 1 = Non-smoker, 2 = Smoker.
FYI, the file originally data entry is in SPSS and i'm convert it to STATA format. When I change the variable with: 0 = Non-smoker, 1 = Smoker
gllamm running smoothly.

Thanks for all response.
 

bukharin

RoboStataRaptor
#9
I also apologise for my late response (I was gallivanting around the globe).

Glad you got it working. It's always nice to post the solution so that future forum searches will be helpful for people with similar problems (see also point 8 here).