[R] Can R do zero inflated gamma regression?

[Ben Bolker]

siriustar <qinlangjinan <at> live.cn> writes:

> 
> Hi, Dear R-help
> I know there are some R package to deal with zero-inflated count data. But I
> am now looking for R package to deal with zero-inflated continuous data.
> 
> The response variable (Y) in my dataset contains a larger mount of zero and
> the Non-zero response are quite right skewed. Now what i am doing is first
> to use a logistic regression on covariates (X) to estimate the probability
> of Y being 0. Then focus on the dataset where Y is not zero, and run a
> linear regression or gamma glm to estimate the association between Y and X
> when Y is not zero.
> However, the linear regression and gamma glm model fit my data poorly.
> 
> So, I am thinking maybe a zero inflated gamma or zero inflated lognormal
> regression are helpful, where I can estimate the probability of Y being zero
> and the association between non zero Y and X at the same time. 
> However, I dont know which R package can do that. 

  I think your 'conditional' strategy is quite useful in general, and
may in general give you the same answers as the zero-inflated approach
you're suggesting.  Perhaps there are some other issues with the
conditional (gamma GLM) parts of your analysis?  Have you tried simple
log-linear regression (i.e. assuming that the non-zero values are
lognormally distributed)?

  I would recommend reading this thread in the r-sig-ecology mailing list:

http://thread.gmane.org/gmane.comp.lang.r.ecology/2124