Dear Alexandre,
Indeed the solution by Florian below is very nice. If you really want to
go the other direction you can use `as.SpatialGridDataFrame.im()` from
the `maptools` package.
By the way, based on what you write, I think you mean to use the value
7.5 (=0.75m) rather than 0.75 (=0.075 m) for your kernel bandwidth.
Furthermore, the argument to set the standard deviation of the
(isotropic Gaussian) smoothing kernel in `density.ppp()` is `sigma`, so
you probably want:
d_pines <- density(swedishpines, sigma = 7.5)
If you like to work in meters rather than decimeters you can rescale the
point pattern:
pines <- rescale(swedishpines, s = 10, unitname = "m")
d_pines <- density(pines, sigma = 0.75)
Regards,
Ege
On 09/27/2018 10:01 AM, Florian Betz wrote:
Dear Alexandre,
instead of converting the image to a SpatialPixelDataFrame converting to
a raster object might be an alternative.
r_pines<-raster(d_pines)
writeRaster(r_pines, "Pines.tif")
Regards,
Florian
Am 26.09.2018 um 22:25 schrieb ASANTOS via R-sig-Geo:
Dear R-sig-geo Members,
I've like to create Kernel image results as *tif using an object of
density() function output in spatstat package. But in my example,
doesn't work when I try:
#Packages
library(spatstat)
library(raster)
library(rgdal)
#Swedishpines's data set in spatstat package
data(swedishpines)
plot(swedishpines)
#CSR with K-Ripley test
csr_pines <- envelope(swedishpines, Kest, nsim=99)
plot(csr_pines)
# r=0.75 is outside CSR
#Kernel representation using 0.75 as bandwidth
d_pines<-density(swedishpines, bw=0.75)
plot(d_pines)
#Create TIFF image
r_pines <- as(d_pines, "SpatialPixelsDataFrame")
writeGDAL(r_pines, "Pines.tif")
#
r_pines <- as(d_pines, "SpatialPixelsDataFrame") Error in as(d_pines,
"SpatialPixelsDataFrame") : no method or default for coercing “im” to
“SpatialPixelsDataFrame”
Please any ideas for corrected this?
Thanks in advanced,
Alexandre
_______________________________________________
R-sig-Geo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
