"0 (empty)" during logistic regression

[Mira Syadza]

Dear Forum,

I am having a problem with a logistic regression that uses an interaction variable, dummy variables.

eststo : logit miskin av_soccap totsc_sq jenkel tanggungan tinggal i.rumah i.Pendd

note: 6.Pendd != 0 predicts failure perfectly
6.Pendd dropped and 1 obs not used

Iteration 0: log likelihood = -489.11406
Iteration 1: log likelihood = -315.05884
Iteration 2: log likelihood = -263.62869
Iteration 3: log likelihood = -257.65542
Iteration 4: log likelihood = -257.57041
Iteration 5: log likelihood = -257.57006
Iteration 6: log likelihood = -257.57006

I have problem 0 (empty) logistic regression in S3
what should i do to this problem? I don't understand why it can be 0 (empty). Please, help me

[Rich Goldstein]

the message from Stata is telling you that 6.Pendd (presumably S3 based on what you say) has only one per and therefore tells you nothing about what is going on and is dropped by Stata because it is, statistically at least, non-informative; some of the other message are not clear because they appear to be from the add-on package; it may be easier, in the future, to answer if you just give the basic model information

[Mira Syadza]

Dear Rich Goldstein
Thank you very much for you replies. Sir, can you explain in more detail. I still can't understand it, how do I interpretation the 0 (empty) variable. Please, help me