I have a dataframe consisting of the location of several individuals, each identified by an ID, and the corresponding values of some environmental variables, all continuous except for one that is factorial. I have created a glmer model and now I want to create a prediction map using this model and the variables.
The summary of my dataframe is as follow:
> summary(jackal.Data.rsf)
X Y Used ID elevation
Min. :4548927 Min. :2524750 Mode :logical Alberto : 7733 Min. : -0.5151
1st Qu.:4570832 1st Qu.:2538306 FALSE:43960 Isabella: 6809 1st Qu.: 95.2394
Median :4583158 Median :2555639 TRUE :4396 Pepe : 4114 Median : 172.2590
Mean :4585547 Mean :2555266 Sergio : 2167 Mean : 289.8413
3rd Qu.:4597305 3rd Qu.:2571271 Torre :15070 3rd Qu.: 354.5094
Max. :4635669 Max. :2591750 Trilli :10923 Max. :1869.0605
Yama : 1540
slope land_cover dist_water dist_road
Min. : 0.0 1 : 4293 Min. : 0.000 Min. : 0.0
1st Qu.:242.0 2 :20206 1st Qu.: 7.071 1st Qu.: 100.0
Median :249.0 3 :14331 Median : 16.155 Median : 200.0
Mean :242.2 4 : 8583 Mean : 21.367 Mean : 358.5
3rd Qu.:250.0 5 : 754 3rd Qu.: 31.048 3rd Qu.: 447.2
Max. :250.0 7 : 171 Max. :123.004 Max. :4167.7
NA's: 18
The model is created using the function
> m2.glmer <- glmer (Used ~ scale(elevation) + scale(slope) + land_cover
+ + scale(dist_water) + scale(dist_road) + (1|ID),
+ data=jackal.Data.rsf, family = binomial(link="logit"),
+ glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 100000)))
I then stack the rasters
> rasters <- stack(elevation, slope, land_cover, dist_water, dist_road)
But when I try to create the prediction map I get the error:
> predictions = raster::predict(object = rasters, model=m2.glmer, fun=predict)
Error in `contrasts<-`(`*tmp*`, value = contrasts.arg[[nn]]) :
contrasts apply only to factors
I have tried to look for the solution to this error but I can only find solutions regarding this error when it appears in the creation of the model, which is not my case.