Gamma log link fitting to a GLM

Discuss statistics related things

by sci_djw » Sun Sep 22, 2019 1:54 pm

Hi there,

I'm new to jamovi and I think this is a fantastic piece of software!

I have a rather complex longitudinal dataset where I want to examine the effects of drugs and radiation on cell death over time. I also have multiple replicates and wish to include these either as a repeated measure or random factor. As far as I'm concerned, I'm just wanting to focus on the interaction between drug and radiation and its effects on cell death (Dependent variable), but wish to consider both time and replicate number when computing the analysis.

I performed glmer in R where I ran a gamma log link as the residuals of my linear model (lme4) where not normal. However, I'm having trouble trying to run this a similar model in jamovi. I was wondering if something could help/advise me on how to perform a gamma log link correction to my data.

Kind regards,

Dave
sci_djw
 
Posts: 1
Joined: Sun Sep 22, 2019 1:40 pm

by jonathon » Sun Sep 22, 2019 9:37 pm

hi dave,

see if the GAMLj module serves your needs.

cheers

jonathon
User avatar
jonathon
 
Posts: 933
Joined: Fri Jan 27, 2017 10:04 am

by mcfanda@gmail.com » Tue Sep 24, 2019 2:58 pm

Hi Dave

the upcoming GAMLj new version (1.6.0), due in a week, will allow to run generalized linear models with any combination of distribution and link function known to stats::family(), including Gamma(log). If you need it now, you can find the code here https://github.com/gamlj/gamlj/tree/Version.1.6.0
mcfanda@gmail.com
 
Posts: 145
Joined: Thu Mar 23, 2017 9:24 pm


Return to Statistics