[R-sig-Geo] raster grid does not fit to SpatialLines and -lPolygons plotted overlays

[Roger Bivand]

On Fri, 19 Mar 2010, Axel Hille wrote:

> Dear OSGEO list members,
>
> I try it again hoping that someone could give me some hint.
> I have the problem that a raster grid does not fit to two additional
> spatial objects in the coordinate space of the plot, when plotted as
> overlay.
> I presume I need to recenter the raster grid but don`t know how? Any help
> is welcome.

Don't presume, it is usually misleading.

You have not motivated this or your previous question, and are writing 
from a German free email provider, so we have no idea whether this is 
homework or what. Remember that a concise motivation and an informative 
affiliation assist helpers in perceiving "where you are coming from". The 
reason for no reply to your earlier question was most likely that you are 
not clear enough (or concise enough - most of the printout you included 
was irrelevant and obscures the issue).

Most likely the coordinate reference systems of the data differ, but I 
don't know if you know what a coordinate reference system is - a graduate 
geography student would, and wouldn't ask this question.

Your workflow (point data to Geneland, grid interpolation in a "haplo" of 
unknown provenance, Arc ASCII grid read in) leaves lots of possibilities 
for the raster to get out of register with respect to the (assumed) UTM32 
points. Do you know definitely that the boundary polygons and rivers are 
also in UTM32? They are declared as such, but are the declarations 
correct? Do the input points plot correctly (are they in UTM32?)? Have you 
exported the input points/rivers/boundaries to KML to double-check in GE?

Try interpolating with some other tool and see whether the same problem 
occurs, or whether it is created in Geneland/haplo.

Hope this helps,

Roger

> Cheers
> Axel Hille
>
> Infos to problem:
> First, I exported the "Geneland" (R-package) output
> file (population assignments, georeferenced ) , where the coordinates of
> the individual samples are given in meters in the UTM-32N projection
> (Middle Europe)
> R-2.10.1-code:
> < head(mat.neu)
> <??? UTM.32E UTM.32N
> < 1 555799.2 5538765
> < 2 555825.5 5538792
>
> by means of the function "haplo" by Cristoph Heibl as an ARCView ascii
> raster grid
>
> R-2.10.1 "file.show":
> < gncols 100
> < nrows 100
> < xllcorner -276050.539270941
> < yllcorner 4802126.61354771
> < cellsize 14070.0737708024
> < NODATA_value -9999
> < 0.3275 0.3225 0.315 0.325 ..........
>
> With readAsciGrid of "maptools" I read in the file,
>
> R-2.10.1-code:
> < (file1 <-? paste(MCMC.PATH.TRANS3,"trans3.map.arcgis1",sep=""))
> < x.grid <- readAsciiGrid(file1, as.image=FALSE, proj4string=
> CRS("+init=epsg:32632"))
> < summary(x.grid)
> < Object of class SpatialGridDataFrame
> < Coordinates:
> <???????????????? min???? max
> < coords.x1 -276050.5 1130957
> < coords.x2 4802126.6 6209134
> < Is projected: TRUE
> < proj4string :
> < [+init=epsg:32632 +proj=utm +zone=32 +ellps=WGS84 +datum=WGS84 +units=m
> +no_defs +towgs84=0,0,0]
> < Number of points: 2
> < Grid attributes:
> <?? cellcentre.offset cellsize cells.dim
> < 1???????? -269015.5 14070.07?????? 100
> < 2???????? 4809161.7 14070.07?????? 100
> < Data attributes:
> <??? Min. 1st Qu.? Median??? Mean 3rd Qu.??? Max.
> <? 0.0000? 0.3187? 0.3325? 0.3427? 0.3463? 1.0000
>
> Then I wanted to plot it and add two additional Spatial objects, a
> SpatialPolygonsDataFrame (European map in the same projection)
> and a SpatialLine object (rivers),
>
> R-2.10.1-code:
> str(eur.SPpdf2, max.level=2)
>> Formal class 'SpatialPolygonsDataFrame' [package "sp"] with 5 slots
>> ?? ..@ data?????? :'data.frame': 66 obs. of? 20 variables:
>> ?? .. ..- attr(*, "data_types")= chr [1:20] "C" "C" "C" "C" ...
>> ?? ..@ polygons?? :List of 66
>> ?? ..@ plotOrder? : int [1:66] 12 47 44 59 38 33 11 34 1 5 ...
>> ?? ..@ bbox?????? : num [1:2, 1:2] -1112905 4041472 1351429
> 6597694
>> ?? .. ..- attr(*, "dimnames")=List of 2
>> ?? ..@ proj4string:Formal class 'CRS' [package "sp"] with 1 slots
>
> R-2.10.1-code:
> summary(eur.SPpdf2)
>> Object of class SpatialPolygonsDataFrame
>> Coordinates:
>> ??????? min???? max
>> x -1112905 1351429
>> y? 4041472 6597694
>> Is projected: TRUE
>> proj4string :
>> [+init=epsg:32632 +proj=utm +zone=32 +ellps=WGS84 +datum=WGS84 +units=m
> +no_defs +towgs84=0,0,0]
>> Data attributes:???? LAYER??? FIPS_ADMIN GMI_ADMIN .......
>
> R-2.10.1-code:
> eur.river.SPldf2 <- spTransform(eur.river.SPldf,
> CRS("+init=epsg:32632"))?? # UTM32N
> bb.river <- eur.river.SPldf2 at bbox
> bb.river
>> ????????? min?????? max
>> r1? 298399.7? 849034.5
>> r2 5265325.5 5749782.6
> ?
> The plot is ok for the vector shapes, but the raster does not align within
> the coordinate space.
> As I'm not so familiar with the information given for the raster file I#m
> asking the rgeo users
> what to do to place the raster correctly within the plot.
> Raster infos:
> object.size(x.grid)
>> [1] 82916
> object.size(slot(x.grid, "grid"))
>> [1] 556
> object.size(slot(x.grid, "grid.index"))
>> [1] 24
> object.size(slot(x.grid, "coords"))
>> [1] 344
>
> object.size(slot(x.grid, "bbox"))
>> [1] 440
> ?
> The plot is ok for the vector shapes, but the raster does not align within
> the coordinate space.
> As I'm not so familiar with the information given for the raster file I#m
> asking the rgeo users
> what to do to place the raster correctly within the plot.
> Raster infos:
> object.size(x.grid)
>> [1] 82916
> object.size(slot(x.grid, "grid"))
>> [1] 556
> object.size(slot(x.grid, "grid.index"))
>> [1] 24
> object.size(slot(x.grid, "coords"))
>> [1] 344
>
> object.size(slot(x.grid, "bbox"))
>> [1] 440
>
> x.grid at bbox
>> ???????????????? min???? max
>> coords.x1 -276050.5 1130957
>> coords.x2 4802126.6 6209134
>> x.grid at bbox["coords.x1", ]
>> ?????? min?????? max
>> -276050.5 1130956.8
>
>> print(slot(slot(x.grid, "grid"), "cellsize"), digits=20)
> [1] 14070.0737708024 14070.0737708024
>
>> print(slot(slot(x.grid, "grid"), "cellcentre.offset"), digits=20)
> [1] -269015.50238554 4809161.65043311
>
>> print(slot(slot(x.grid, "grid"), "cells.dim"), digits=20)
> [1] 100 100
>>
>
>        x.grid at bbox["coords.x1", ]
>> ?????? min?????? max
>> -276050.5 1130956.8
>
>> print(slot(slot(x.grid, "grid"), "cellsize"), digits=20)
> [1] 14070.0737708024 14070.0737708024
>
>> print(slot(slot(x.grid, "grid"), "cellcentre.offset"), digits=20)
> [1] -269015.50238554 4809161.65043311
>
>> print(slot(slot(x.grid, "grid"), "cells.dim"), digits=20)
> [1] 100 100
>>
> -- GMX DSL: Internet, Telefon und Entertainment f?r nur 19,99 EUR/mtl.!
>        http://portal.gmx.net/de/go/dsl02
>

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no