rsf.select(): Error in eigen(sigma)

[Philipp Maleko]

Hello friends,

I am running rsf.select on birds using a list of 5 factorized landcover classes. While this works for most birds, for one particular individual I keep getting the error "Error in eigen(sigma) : infinite or missing values in 'x'". Interestingly, this error occurs after going through 20 iterations of other landcover class combinations. In the past, i had this issue when attempting rsf.select() on a raster that was massive, and corrected it by reducing the extent. This time I get this error under any raster extent, big or small. Although my sample of points is low (n = 53) my AKDE DOF is okay (41)... Curious if anyone else had this issue and can suggest any solutions.

Thank you so much in advance!!

Philipp

[Sonni B]

Hi Phillipp,

Unfortunately, I don't have a solution for you, but I'm adding my own experience to this thread in case it's helpful.  I am also getting the same error, and rerunning the code will also give me - " Error in rsf.fit(...): NA values in sampled variables x,y,rr at points 1,2,3,4,5,6,.., although I've tripled checked that none of my rasters, telemetry objects, or AKDEs have NA values. The weirdest part is that I've run this exact code before and did not have this issue. 

I think my data are similar to yours, Phillipp, in that I have 4 binary rasters of interest cropped to the extent of each individual telemetry object. Maybe we are both overlooking something? Sorry I can't be of more help.

Best,

Sonni

[Sonni B]

Hi Phillipp,

I was able to get my code to run last night! Here's what I did:

First, I made sure all my telemetry objects were using a 'tpeqd' projection. Doing 'projection(DATA) <- median(DATA)' would convert the projection to 'aeqd', so I got rid of that code. All my animals were using the same CRS even without that.

Second, I was using a for loop to run rsf.select() for each individual, and I would get the error on the 6th animal. All I did was continue running the for loop from where it last left off. Seems like it was a RAM issue. 

Your code also seems pretty computationally intense, so maybe trying to break it down into smaller chunks may help? 

I hope this helps!

Sonni