Maximilian Pichler

Hm, yes, but I don't know if it makes much sense for hundreds of species - of course we could always limit it to the upper and lower 5% quantile...

Hi, it seems somehow related to the new plotting engine of rstudio. Other ppl encountered the same issue on ubuntu when a) the R version and RStudio version didn't match...

Hi Kara, I assume you installed the GPU drivers via ubuntu's apt package manager? I never use it because the drivers are often completely outdated and rarely match my GPUs....

Hi Kara, sorry but I don't think I can help you with this. If the problem occurs only if you load sjSDM, maybe it is related to the ggplot package...

Hi @thk686, can you please post the output of sessionInfo() here? It seems to be a problem related to ubuntu and the latest R version.

Hi @WevertonBio, Difference between importance and anova: - importance(...) is a variance partitioning based on [Ovaskainen et al., 2017](https://onlinelibrary.wiley.com/doi/full/10.1111/ele.12757) where we separate the variances for the fixed and random effects....

Hi @WevertonBio, the importance(....) can calculate the VP only for the species, not the sites. I made a mistake in #73.

Hi @dansmi-hub, Unfortunately, we haven't implemented a resid/residual method yet. However, you can calculate it yourself. Stand. Pearson residuals: ``` com = simulate_SDM(env = 3L, species = 7L, sites =...

Hi @Zurapiti, The idea is to use the responses summed (across sites) on the linear scale and then see if there is a trend in certain environmental variables: ``` library(sjSDM)...

raw == on the scale of the linear predictor (before the inverse link function) , might be good idea to change it to the same default values as in predict.glm...