[R-sig-Geo] gdistance, shortest path => "cannot derive coordinates from non-numeric matrix"
Jacob van Etten
jacobvanetten at yahoo.com
Fri Aug 12 11:14:48 CEST 2011
Oh, but now it throws a straightforward error: "At structural_properties.c:4277 :Couldn't reach some vertices".
This means that path between A and B does not exist, because they are on different unconnected "islands". Try again, after giving a small positive value to all raster cells that are 0 now.
Something like:
rognes.r[rognes.r<1] <- 1
You don´t have to transform the layer.
Having an NA crs slot in your raster is not a problem in your case, because isLonLat(rognes.r) would give FALSE, which is what you want.
Jacob.
--- On Fri, 12/8/11, Mathieu Rajerison <mathieu.rajerison at gmail.com> wrote:
> From: Mathieu Rajerison <mathieu.rajerison at gmail.com>
> Subject: Re: [R-sig-Geo] gdistance, shortest path => "cannot derive coordinates from non-numeric matrix"
> To: r-sig-geo at r-project.org
> Date: Friday, 12 August, 2011, 9:17
> The problem seems to come from my
> raster values.
>
> 1) With all values equal to 1, as in your example, it
> works
>
> > rognes.r
> class : RasterLayer
> dimensions : 389, 363, 141207 (nrow, ncol, ncell)
> resolution : 25, 25 (x, y)
> extent : 885426, 894501, 6283406, 6293131 (xmin, xmax,
> ymin, ymax)
> coord. ref. : +proj=utm +zone=31
> values : in memory
> min value : 1
> max value : 1
> >(...)
> OK, the path is calculated
>
>
> 2) But with my original raster, it doesn't.
> > rognes.r
> class : RasterLayer
> dimensions : 389, 363, 141207 (nrow, ncol, ncell)
> resolution : 25, 25 (x, y)
> extent : 885426, 894501, 6283406, 6293131 (xmin, xmax,
> ymin, ymax)
> coord. ref. : +proj=utm +zone=31
> values : in memory
> min value : 0
> max value : 100
>
> > summary(rognes.r)
> Cells: 141207
> NAs : 0
> Min. 0.00
> 1st Qu. 0.00
> Median 0.00
> Mean 16.03
> 3rd Qu. 0.00
> Max. 100.00
>
> > rognes.t = transition(rognes.r, transitionFunction =
> mean, directions=8)
> > A = c(887543,6286048)
> > B = c(890911,6289681)
> > path = shortestPath(rognes.t, A, B,
> output="SpatialLines")
> Erreur dans .local(obj, ...) :
> cannot derive coordinates from non-numeric matrix
> De plus : Message d'avis :
> In get.shortest.paths(adjacencyGraph, indexOrigin,
> indexGoal) :
> At structural_properties.c:4277 :Couldn't reach some
> vertices
>
> Do you see any problem with my raster values?
>
> 2011/8/12 Mathieu Rajerison <mathieu.rajerison at gmail.com>
>
> > I've not specified crs reference when I created my
> raster which is in
> > EPSG:2154. Is it a problem?
> >
> > Do I have to reproject it from EPSG:2154 to Lon/Lat
> before using
> > TransitionLayer calculations?
> >
> >
> > My sessionInfo() is as follows:
> > R version 2.13.1 (2011-07-08)
> > Platform: i386-pc-mingw32/i386 (32-bit)
> >
> > locale:
> > [1] LC_COLLATE=French_France.1252
> LC_CTYPE=French_France.1252
> > [3] LC_MONETARY=French_France.1252 LC_NUMERIC=C
> > [5] LC_TIME=French_France.1252
> >
> > attached base packages:
> > [1] stats graphics grDevices utils datasets methods
> base
> >
> > other attached packages:
> > [1] maptools_0.8-9 foreign_0.8-44 gdistance_1.1-1
> > [4] Matrix_0.999375-50 lattice_0.19-30 igraph_0.5.5-2
> > [7] spatstat_1.22-3 RandomFields_2.0.45 deldir_0.0-14
> > [10] mgcv_1.7-6 raster_1.8-39 sp_0.9-83
> >
> > loaded via a namespace (and not attached):
> > [1] grid_2.13.1 nlme_3.1-101 tools_2.13.1
> >
> >
> > Thanks for the quick answer!
> >
> > 2011/8/11 Jacob van Etten <jacobvanetten at yahoo.com>
> >
> >> Dear Mathieu,
> >>
> >> From what I can see, the error is not due to the
> TransitionLayer, but is
> >> about the coordinates.
> >>
> >> The following code works with the versions of
> gdistance and raster that
> >> are currently on CRAN (gdistance_1.1-1,
> raster_1.9-5):
> >>
> >> library(gdistance)
> >> rognes.r <- raster(nrows=50, ncols=50,
> xmn=887000, xmx=892000,
> >> ymn=6285000, ymx=6290000, crs="+proj=utm
> +zone=31")
> >> rognes.r[] <- 1
> >> rognes.t = transition(rognes.r, transitionFunction
> = mean, directions=8)
> >> A = c(887543,6286048)
> >> B = c(890911,6289681)
> >> path = shortestPath(rognes.t, A, B,
> output="SpatialLines")
> >>
> >> Next time, provide the results of sessionInfo()
> when you post to this
> >> list. Also, try to provide reproducible code (like
> in my example). With the
> >> information you give now, I can only guess the
> cause of your problem.
> >>
> >> Jacob
> >>
> >> --- On Thu, 11/8/11, Mathieu Rajerison <mathieu.rajerison at gmail.com>
> >> wrote:
> >>
> >> > From: Mathieu Rajerison <mathieu.rajerison at gmail.com>
> >> > Subject: [R-sig-Geo] gdistance, shortest path
> => "cannot derive
> >> coordinates from non-numeric matrix"
> >> > To: r-sig-geo at r-project.org
> >> > Date: Thursday, 11 August, 2011, 17:59
> >> > Hi,
> >> >
> >> >
> >> > I'm using gdistance to determine
> optimal wildlife
> >> > corridors. My transition
> >> > layer's name is rognes.t, rognes being a
> city name.
> >> >
> >> > I want to determine a corridor between two
> major ecologic
> >> > niches with
> >> > centroids coordinates A and B.
> >> >
> >> > But when launching shortestPath, i get the
> following
> >> > error:
> >> > > rognes.t = transition(rognes.r,
> transitionFunction =
> >> > mean, directions=8)
> >> > > A = c(887543,6286048)
> >> > > B = c(890911,6289681)
> >> > > path = shortestPath(rognes.t, A, B,
> >> > output="SpatialLines")
> >> > Erreur dans .local(obj, ...) :
> >> > cannot derive coordinates from non-numeric
> matrix
> >> > De plus : Message d'avis :
> >> > In get.shortest.paths(adjacencyGraph,
> indexOrigin,
> >> > indexGoal) :
> >> > At structural_properties.c:42
> >> >
> >> > My transitionMatrix has these
> characteristics:
> >> > > min(transitionMatrix(rognes.t));
> >> > max(transitionMatrix(rognes.t));
> >> > mean(transitionMatrix(rognes.t))
> >> > [1] 0
> >> > [1] 100
> >> > [1] 0.0009044702
> >> >
> >> > All values are numeric..
> >> >
> >> >
> >> > Any help would be greatly appreciated,
> >> >
> >> > mathieu
> >> >
> >> > [[alternative HTML
> version deleted]]
> >> >
> >> >
> _______________________________________________
> >> > R-sig-Geo mailing list
> >> > R-sig-Geo at r-project.org
> >> > https://stat.ethz.ch/mailman/listinfo/r-sig-geo
> >> >
> >>
> >
> >
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>
More information about the R-sig-Geo
mailing list