julian-wittische
commented
2 years ago The issue was associated with a mistake when setting the CRS of the sample locations. Using the following led to a successful run:
samples <- read.table("EGK_SampleLoc.txt", sep="") sp.dat_Julian <- SpatialPoints(samples[,c(1,2)], proj4string = CRS("+proj=longlat +datum=WGS84")) sp.dat_Julian_reproj <- spTransform(sp.dat, CRS("+proj=utm +zone=56 +south"))
wpeterman
G'day,
I've been trying to optimise a singular surface through the Julia Circuitscape function, but have been running into unexpected results within my optimisation. Whenever I run my optimisation, I'm met with a constant mean and best value of '-9999' for each iteration. As well, my RStudio continuously crashes after 23 iterations of that value. My area of interest is quite irregular in shape - so I expected it to have a fair amount of 'nodata' values in the background mask/matrix. However, after I examined the .asc file I created, I noticed that there were no values of '-9999' but rather all of those values were transformed to '0'. I thought it might be an issue with circuitscape originally, but when I tried it with 'commuteDistance' I got a similar output -
Error in { : task 3 failed - "0 (non-NA) cases"That being said, I still have these questions:
I've supplied the data that I'm using through the link below. Any information on this will be very much appreciated! I feel like this is a simple fix, and has something to do with the raster data that I'm using, but for the life of me I can't figure it out.
Thanks for your time! Chris
Data: https://drive.google.com/drive/folders/17P0zNoBj4x4T1DeGA36U3PpnMcBhDR-i?usp=sharing Code: `
Import Library
library(ResistanceGA)
SINGLE SURFACE OPTIMISATION
Create Directory for Examples
if("ResistanceGA_Examples"%in%dir("D:/ResistanceGA/R/V1_Output")==FALSE) dir.create(file.path("D:/ResistanceGA/R/V1_Output", "V1_Test"))
Create Subdirectory for Single Surface
dir.create(file.path("D:/ResistanceGA/R/V1_Output/V1_Test", "SingleSurface"))
Create Directory Object for .asc Files and results
write.dir <- "D:/ResistanceGA/R/V1_Output/V1_Test/SingleSurface/"
Set working directory
setwd("D:\ResistanceGA\Data\ASC")
Optimise using Julia's Circuitscape
Sample locations
samples <- read.table("D:\ResistanceGA\EGK_SampleLoc.txt", sep="") sp.dat <- SpatialPoints(samples[,c(1,2)]) crs(sp.dat) <- CRS("+proj=utm +zone=56 +south") plot(sp.dat, pch = 19)
Import GeoTIFF and write to .asc in write.dir
msavi <- raster("MSAVI2_Test.tif") plot(sp.dat, pch = 19, col = "blue") plot(msavi) writeRaster(msavi, filename = paste0(write.dir, "MSAVI2_Test.asc"), overwrite = TRUE)
Genetic distance measured between sample locations (chord distance)
gen_dist2 <- read.table('D:\ResistanceGA\R\V1_Output\V1_Test\SingleSurface\fst.txt', sep="", header = FALSE)
Specify bin Directory and connect Julia to R
JULIA_HOME <- "C:/Users/User/AppData/Local/Programs/Julia-1.6.3/bin/" JuliaCall::julia_setup(JULIA_HOME)
Run Prep Functions
GA.inputs <- GA.prep(ASCII.dir = write.dir, Results.dir = "D:\ResistanceGA\R\V1_Output\V1_Test\SingleSurface\", parallel = 4)
jl.inputs <- jl.prep(n.Pops = length(sp.dat), response = lower(as.matrix(gen_dist2)), cholmod = TRUE, precision = FALSE, CS_Point.File = sp.dat, JULIA_HOME = JULIA_HOME)
Run Optimisation
jl.optim <- SS_optim(jl.inputs = jl.inputs, GA.inputs = GA.inputs)`
Output: `
Test Passed