Michelle Ensbey
2009-Oct-22 07:39 UTC
[R] Predictions and prediction intervals for model.averaging dRedging::model.avg
Hi R all Does anyone know how to get predictions and/or put prediction intervals on model averaged results from dRedging::model.avg for a poisson GLM. Predict doesn't work in this case like it does dirently on a glm?> glm_aab_null<- glm(aabundance~1, family=poisson, data=tenuredata).....> glm_aab_rain_rock_tenure<- glm(aabundance~rainfall + rock + tenure, family=poisson, data=tenuredata)> modelavg_aab_all<-model.avg(glm_aab_null,glm_aab_rain,glm_aab_rock,glm_aab_tenure,glm_aab_rain_rock,glm_aab_rain_tenure,glm_aab_rock_tenure,glm_aab_rain_rock_tenure)> predict_tenure<-predict(modelavg_aab_all,se.fit=T,data.frame(rock=60,rainfall=1000,tenure=levels(tenure)), type = "response")Error in UseMethod("predict") : no applicable method for "predict" I imagine people would like to predict from their model averaging results all the time (with confidence), so please let me know if there is a function out there or someone has some adaptable code to do this. Cheers Miki R version 2.9 --------------------------&:-)>---------------------------------- Michelle (Miki) Ensbey michelle.ensbey@nt.gov.au [[alternative HTML version deleted]]