search

biodiverse / spOccupancy

Single-species, Multi-species, and Integrated Spatial Occupancy Models
https://www.jeffdoser.com/files/spoccupancy-web/
GNU General Public License v3.0
52 stars 8 forks source link

.doSnowGlobals not found #32

Closed swbigelow closed 1 year ago

swbigelow commented 1 year ago

I am working through the Hubbard Brook tutorial. When I run spPFOcc(), I get error "Error in checkForRemoteErrors(lapply(cl, recvResult)) : one node produced an error: object '.doSnowGlobals' not found". I understand from StackOverflow that this is a parallel computing issue-somehow the cores are not getting the information they need. I have tried installing 'doSNOW' and 'doParallel' packages, and various ways of passing the location of my packages to the computer. I should note that occasionally I get a message from Windows"(Windows firewall defender has blocked some features of R for Windows GUI front-end on all public and private networks". Do you have any suggestions?

doserjef commented 1 year ago

Hi @swbigelow,

Thanks for the note. I'm not very familiar with parallelization on Windows systems, but it does seem like this could be happening as a result of the windows firewall on your machine (i.e., take a look at this guide). Are you running spPGOcc() with cross-validation (i.e., are you specifying the k.fold argument?). Cross-validation is the only time that that the doParallel package is used, so I'm thinking that you should be able to run the functions without specifying that argument.

Thanks,

Jeff

swbigelow commented 1 year ago

I just ran spPGOcc() without cross-validation (k-fold argument) and it ran like a charm, no more '.doSnowGlobals' error message. I have now Bayesed! It's exciting. Thanks very much for the troubleshooting help. I look forward to running this package for black rail and Bachmann's sparrow.

doserjef commented 1 year ago

Awesome! Glad to hear you got it to run!