Hello all,
Has anyone come across this error "CHOLMOD factorization was
unsuccessful" when using the resistance distance function in the new
package gdistance? I appreciate the help, this forum is really allowing me
to get past these issues.
#Read in the coordinates
> head(cord)
coords.x1 coords.x2
0 -73.36111 6.810003
1 -73.45474 6.464962
2 -76.28194 1.137222
3 -76.40000 1.113333
4 -76.53278 1.233056
5 -76.27610 1.600061
> class(cord)
[1] "matrix"
#Read in the elevation raster
elevr<-raster("D:/Ben/GIS/Data/DimDiv/StudyArea_1km")
>Elevr
class : RasterLayer
dimensions : 2047, 1648, 3373456 (nrow, ncol, ncell)
resolution : 0.007821463, 0.007821462 (x, y)
extent : -80.82072, -67.93095, -4.788296, 11.22224 (xmin, xmax, ymin,
ymax)
coord. ref. : +proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs
+towgs84=0,0,0
values : D:\Ben\GIS\Data\DimDiv\StudyArea_1km
min value : 1
max value : 6169
#create a friction surface, i've also tried directions = 4
tmean<-transition(elevr, transitionFunction = mean, directions=8)
> tmean
class : TransitionLayer
dimensions : 2047, 1648, 3373456 (nrow, ncol, ncell)
resolution : 0.007821463, 0.007821462 (x, y)
extent : -80.82072, -67.93095, -4.788296, 11.22224 (xmin, xmax, ymin,
ymax)
coord. ref. : +proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs
+towgs84=0,0,0
values : conductance
matrix class: dsCMatrix
#perform the geocorrection (i've also tried skipping this step)
resistancesurface <- geoCorrection(tmean, type = "r", multpl = FALSE)
> resistancesurface
class : TransitionLayer
dimensions : 2047, 1648, 3373456 (nrow, ncol, ncell)
resolution : 0.007821463, 0.007821462 (x, y)
extent : -80.82072, -67.93095, -4.788296, 11.22224 (xmin, xmax, ymin,
ymax)
coord. ref. : +proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs
+towgs84=0,0,0
values : conductance
matrix class: dsCMatrix
#Get matrix of resistance distances.
> randomwalk<-resistanceDistance(resistancesurface,cord)
Error in Cholesky(Lr) : CHOLMOD factorization was unsuccessful
In addition: Warning message:
In Cholesky(Lr) :
Cholmod warning 'not positive definite' at
file:../Cholesky/t_cholmod_rowfac.c, line 432
>
It's a bit difficult to parse. This process works perfectly for the
costDistance function, which is much prettier than the one i wrote in
python.
And while i'm at it, has anyone had a chance to create a matrix of pairwise
shortest paths? That is the distance of the least cost path from every
combination of sites. My first thought is:
pathloop<- function(coord,origin) sapply(coord, function(x)
shortestPath(meansurface, coord[origin], coord[x,], output="SpatialLines")
and then loop that through a 2nd apply loop for each origin point.
As always, i appreciate the help.
Ben Weinstein
--
Ben Weinstein
Graduate Student
Ecology and Evolution
Stony Brook University
http://life.bio.sunysb.edu/~bweinste/index.html
[[alternative HTML version deleted]]
_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
