Dear Peter:

I notice there is a R code for a Zero-inflated Poisson/NB process on the Stanford Political Science Computational Lab (Prof. Simon Jackman) web page. If I were wanting to do a one-inflated model, I would start with that because, at least to my eye, it is very easy to follow. Mind you, I did not try this myself, but I bet you could make it go. In the file zeroinfl.r, look at the function:

zeroinflNegBin <- function(parms){

it is pretty clear you'd have to supply a probability model for the outcomes valued 1 and then fit them into the overall likelihood.

pj


http://pscl.stanford.edu/content.html Peter Flom wrote:

Hello

I am interested in Poisson or (ideally) Negative Binomial regression
with an inflated number of 1  responses

I have seen JK Lindsey's fmr function in the gnlm library, which fits
zero inflated Poisson (ZIP) or zero inflated negative binomial
regression, but the help file states that for ' Poisson or related
distributions  the mixture involves the zero category'.

I had thought of perhaps subtracting 1 from all the counts and then
fitting the ZIP or ZINB models, and then adding 1, but am not sure if
this is legitimate, or if there is some better method.

Contextual details:
The dependent variable is number of primary sexual partners in the last
year.  The independent variables include a) Being married or in a
committed relationship  b) using hard drugs  c) sex  d) age

N is c. 500

Not surprisingly, there are a large number of 1 responses, especially
for those who are married or in a relationship.  More surprisingly, the
mean number of partners is the same (1.05 vs. 1.02) for people in and
not in relationships, but the variances are very different, mostly
because those in a relationhsip are much more likely to say exactly 1.

Thanks in advance

Peter

Peter L. Flom, PhD
Assistant Director, Statistics and Data Analysis Core
Center for Drug Use and HIV Research



--
Paul E. Johnson email: [EMAIL PROTECTED]
Dept. of Political Science http://lark.cc.ku.edu/~pauljohn
1541 Lilac Lane, Rm 504 University of Kansas Office: (785) 864-9086
Lawrence, Kansas 66044-3177 FAX: (785) 864-5700


______________________________________________
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html

Reply via email to