[R-sig-Geo] writeRaster - netcdf - Rotation

[Philipp Reiter]

Hi Robert,
 
sure, m x n can be represented by n x m, but I was confused that after reimporting data as nc I have to transpose and rasterize it first before plotting. Because normally I don't have to do any of those two.
 
 
When I plot the examples from you,
 
image(tmp)
equals
image(writeRaster(tmp, filename='export.nc', format="CDF", overwrite=TRUE, xname='easting', yname='northing'))
and also equals
image(raster('export.nc'))
 
but
image(get.var.ncdf(nc,"layer")) doesn't and I have to transpose and rasterize it first:
image(t(raster(get.var.ncdf(nc,"layer"))))  # Now the plot is correct
 
The last plot is what got me confused. Maybe u can tell me why in this case the transposition and rasterize is needed?
 
Regards,
Philipp
 
 
 
Philipp,

A matrix of dimensions n x m can be represented as m x n, and my example
shows that the data are clearly equivalent (please build on this example if
you have further queries; it is not very helpful to refer to data we cannot
see):

library(raster)
tmp<- raster(nr=5, nc=10,crs='+proj=utm +zone=12')
tmp[] <- 1:ncell(tmp)
r <- writeRaster(tmp, filename='export.nc', format="CDF", overwrite=TRUE,
xname='easting', yname='northing')
all(values(tmp) == values(r))

nc <- open.ncdf("export.nc")
get.var.ncdf(nc,"easting")
get.var.ncdf(nc,"layer")

# result is the same as:
r <- raster('export.nc')
t(as.matrix(r))


Robert




On Mon, Mar 25, 2013 at 3:55 AM, Philipp Reiter <amock at gmx.li> wrote:

> Hi Robert and Pascal,
>
> thanks for the tips, but my problem isn't completely solved. The explicit
> laebling of x and y works, but doesn't change the rotation.
> I followed Pascals hint:
> My raster*-Object has the dimensons: "220 171 x".
> When I export this raster*-Obejct as posted below as nc-file and reimport
> it with the ncdf-Package, the dimensions are: "171 220 x". The explicit
> labeling doesn't change that. The first dimension is always 171 in the
> nc-file, no matter if its names 'easting' or 'northing'. To me it looks
> like raster and ncdf interpret x and y differently?!
>
> Beste regards
> Philipp
>
>
> Philipp,
>
> You are right, the labels easting and northing are reversed. I will fix
> that. You can set them yourself as illustrated below:
>
> library(raster)
> tmp<- raster(nr=5, nc=10,crs='+proj=utm +zone=12')
> tmp[] <- 1:ncell(tmp)
> writeRaster(tmp, filename='export.nc', format="CDF", overwrite=TRUE,
> xname='easting', yname='northing')
> raster('export.nc')
>
> nc <- open.ncdf("export.nc")
> get.var.ncdf(nc,"easting")
>
> Robert
>
> On Fri, Mar 22, 2013 at 7:17 AM, <amock at gmx.li> wrote:
>
> >
> > Hey there,
> > I’ve got a problem concerning exporting a raster*-Object as nc-File:
> > When exporting rasterLayers or rasterBricks as nc-Files with writeRaster
> > from the raster-Package, easting and northing in the resulting nc-File
> are
> > switched, which results in a rotated plot. I can correct this afterwards,
> > but it should be possible to be correct right away?!
> >
> > Example code for a rasterLayer:
> > tmp <- raster(readGDAL(‘example.asc’,p4s="+init=epsg:31467"))
> > writeRaster(tmp, filename=’export.nc’, format="CDF", overwrite=TRUE)
> > nc <- open.ncdf("export.nc")
> > x <- get.var.ncdf(nc,”easting”)
> > y <- get.var.ncdf(nc,”northing”)
> >
> > R version 2.15.3 (2013-03-01)
> > Platform: x86_64-w64-mingw32/x64 (64-bit)
> > ncdf_1.6.6 raster_2.1-16 rgdal_0.8-5 sp_1.0-6
> >
> > I hope someone can give me a hint! Thanks in advance!
> > Philipp
> >
> >
> > _______________________________________________
> > R-sig-Geo mailing list
> > R-sig-Geo at r-project.org
> > https://stat.ethz.ch/mailman/listinfo/r-sig-geo[https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-geo&selection=tfol11b9dc8701907a64][https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-geo&selection=tfol11b9dc8701907a64[https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2F3c.gmx.net%2Fmail%2Fclient%2Fdereferrer%3FredirectUrl%3Dhttps%253A%252F%252Fstat.ethz.ch%252Fmailman%252Flistinfo%252Fr-sig-geo%26selection%3Dtfol11b9dc8701907a64&selection=tfol11b9dc8701907a64]]
> >
> >
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo[https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-geo&selection=tfol11b9dc8701907a64][https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-sig-geo&selection=tfol11b9dc8701907a64[https://3c.gmx.net/mail/client/dereferrer?redirectUrl=https%3A%2F%2F3c.gmx.net%2Fmail%2Fclient%2Fdereferrer%3FredirectUrl%3Dhttps%253A%252F%252Fstat.ethz.ch%252Fmailman%252Flistinfo%252Fr-sig-geo%26selection%3Dtfol11b9dc8701907a64&selection=tfol11b9dc8701907a64]]