I am trying to run a GWR, and using the spgwr package. I just want to run a basic GWR on my data, which looks a little like this:
Concentration | ws | wd | eleva | traffic | dist_to_RR |
---|---|---|---|---|---|
10382 | 2 | 200 | 30 | 11234 | 112.3 |
43722 | 2 | 210 | 30 | 24273 | 2736.3 |
26374 | 2 | 230 | 60 | 23846 | 2163.1 |
etc. etc.
I've converted my dataset to a SpatialPointDataFrame (from an sf) and double checked that all of my variables are numeric. I am trying to use this spatial dataset to run a gwr, and here is the code I have:
band_gwr <- gwr.sel(Concentration~ws+wd+eleva+
traffic+dist_to_RR, data=dat,
gweight=gwr.Gauss, adapt=TRUE)
I can't even get to the next step because this code will time out. It doesn't throw errors, it just runs infinitely long. I have this running on a computing cluster with a lot of cores, and I let it run for 3+ hours and still nothing. I think something must be wrong with the way I've written this, I can't imagine why else this won't work.
I have already checked variable correlation and removed those with Pearson's >0.8, and I also removed all of the variables with >40% homogeneity. There are no NAs in the data. This is an SpatialPointsDataFrame so it does not need a coords argument in the function.
Any ideas on how to go about debugging this?
Thanks!