Hi, I am in the process of producing a quasipoisson model for a dataset I'm working on. The overall GLM works fine but the glmulti function only seems to work with a gaussian model. Here is the code I have at the moment:
#Run first model
model1<-glm(Diadema~., data = X10m, family = quasipoisson(link = log))
summary(model1)
#ANCOVA of all significant variables from model1
model2<-glm(Diadema~Poma*EV*Agraham*Rock, data = X10m, family = quasipoisson(link = log))
summary(model2)
#Model importance (gaussian)
res<-glmulti(Diadema~Poma*EV*Agraham*Rock, data = X10m,
level = 2, fitfunction = glm, crit = "aicc", confsetsize = 100)
print(res)
plot(res)
top<-weightable(res)
top<-top[top$aicc<=min(top$aicc)+2,]
top
summary(res@objects[[1]])
plot(res, type = "s")
eval(metafor:::.glmulti)
coef(res)
mmi<-as.data.frame(coef(res))
mmi<-data.frame(Estimate=mmi$Est, SE=sqrt(mmi$Uncond), Importance=mmi$Importance, row.names = row.names(mmi))
mmi$z<-mmi$Estimate/mmi$SE
mmi$p<-2*pnorm(abs(mmi$z), lower.tail = FALSE)
names(mmi)<-c("Estimate", "Std. Error", "Importance", "z value", "Pr(>|z|)")
mmi$ci.lb<-mmi[[1]]-qnorm(.975)*mmi[[2]]
mmi$ci.ub<-mmi[[1]]+qnorm(.975)*mmi[[2]]
mmi<-mmi[order(mmi$Importance, decreasing = TRUE), c(1,2,4:7,3)]
round(mmi,4)
All of the code works fine, I am just wondering if there was a way to convert the model importance to a quasipoisson model rather than gaussian. Any help would be appreciated.
Also if you need me to convert the data to a public data set, such as iris, let me know.
Cheers.