Solved
New Contributor
Posts: 4

# Glimmix procedure using a weibull distribution

I would like to analyze a multi-location complete randomized block design  where resposne variable is days to maturity (of a crop).  After running a proc severity the weibull distribution was selected.  My questions are :

Is there any information about using the weibull distribution in the glimmix procedure, if so which is the appropriate link function?

Do I need a fully saturated model with weibull distribution  as in a binomial distribution?

Accepted Solutions
Solution
‎12-12-2016 04:47 PM
SAS Super FREQ
Posts: 3,839

## Re: Glimmix procedure using a weibull distribution

I don't think a univariate histogram/analysis of the response variable tells you very much. The DIST= option in GENMOD and GLIMMIX specify the (conditional) error distribution, which might be very different from the distribution of the response. See the pictures in the article "Error distributions and exponential regression models."

To answer your implied question, the GLIMMIX procedure does not support a Weibull distribution function for the errors. It sounds like you are treating the response variable as a continuous variable. If you think you have long-tailed errors, you might look at the gamma or lognormal distributions. (The gamma is sometimes used for waiting times.) Look at the goodness-of-fit statistics and the diagnostic residual plots to guide you in choosing a good model.

All Replies
SAS Super FREQ
Posts: 3,839

## Re: Glimmix procedure using a weibull distribution

You say "after running PROC SEVERITY, a Weibull distribution was selected." Did you use PROC SEVERITY to run a fixed-model regression, or did you just analyze the response variable in isolation (a univariate model)?

New Contributor
Posts: 4

## Re: Glimmix procedure using a weibull distribution

In Isolation (Univariate model)

Solution
‎12-12-2016 04:47 PM
SAS Super FREQ
Posts: 3,839

## Re: Glimmix procedure using a weibull distribution

I don't think a univariate histogram/analysis of the response variable tells you very much. The DIST= option in GENMOD and GLIMMIX specify the (conditional) error distribution, which might be very different from the distribution of the response. See the pictures in the article "Error distributions and exponential regression models."

To answer your implied question, the GLIMMIX procedure does not support a Weibull distribution function for the errors. It sounds like you are treating the response variable as a continuous variable. If you think you have long-tailed errors, you might look at the gamma or lognormal distributions. (The gamma is sometimes used for waiting times.) Look at the goodness-of-fit statistics and the diagnostic residual plots to guide you in choosing a good model.

New Contributor
Posts: 4

## Re: Glimmix procedure using a weibull distribution

Thank you for taking the time  to help Rick.

New Contributor
Posts: 4

## Re: Glimmix procedure using a weibull distribution

Once again thank you Rick,  for pointing me in the right direction.  I will take a look at your sugestions and explore both the lognormal and gamma distribution.

☑ This topic is solved.

Need further help from the community? Please ask a new question.

Discussion stats
• 5 replies
• 206 views
• 2 likes
• 2 in conversation