Hi Etienne,
Just to follow up I took your advice and rewrote my cva analysis using
rasters as opposed to SGDF and the approach has worked well. Your
suggestions in terms of converting 0's to NA was also a winner.
May thanks,
Wesley
On Tue, Feb 5, 2013 at 5:57 PM, Wesley Roberts wrote:
> Many t
Thanks Roger,
That fixed it. I agree with Rainer, it is somewhat non-intuitive.
By the way, I've been using GRASS from within R to process dozens of
images using loops, works like a charm and saved me hours if not days of
repetitive work.
Thanks again and cheers,
Ravi
P.S. The corrected formula
Dear list,
I was trying to write a shapefile within a small function (to calculate
centroids to polygons) but not able to do so
My code was
x<-readOGR(dsn="C:/.", layer="x")
get.centroids<-function(x) {
x.polyset<-SpatialPolygons2PolySet(x)
centroids <- calcCentroid(x.polyset, rollup=2)
d.c
Thanks Rodger,
That fixed it. I agree with Rainer, it is somewhat non-intuitive.
By the way, I've been using GRASS from within R to process dozens of
images using loops, works like a charm and saved me hours if not days of
repetitive work.
Thanks again and cheers,
Ravi
P.S. The corrected formul
Mudrak, Erika [EEOBS] writes:
> Hello r-sig-geo,
> I have a question regarding envelopes for the pair correlation function in
> spatstat.
Questions about the spatstat package should be sent to the package authors.
> The envelopes near r = 0 suddenly stretch to high values. The envelopes
> a
Hi Tanja,
Also, check out the spGLM function in the spBayes package as well as
functions in the geoRglm package.
-Andy
Quoting Juan Tomas Sayago Gomez :
there is also in the package McSpatial the possibility to run
spatial logit
and probit models. But you should check the literature in the
there is also in the package McSpatial the possibility to run spatial logit
and probit models. But you should check the literature in there.
Juan
On Wed, Feb 6, 2013 at 10:09 AM, Milan Sharma wrote:
> Looks like you want autologistic regression.
> Just google "autologistic modelling in R" and you
Another possibility is the blockPredict function in the intamap-package
(also based on gstat), which can do transGaussian kriging.
Fixing lambda = 0 is equal to lognormal kriging.
A new version with a couple of inconsistencies removed and a better
example for lognormal kriging (1.3-18) has just
Looks like you want autologistic regression.
Just google "autologistic modelling in R" and you will get archieves of
previous sessions.
Best,
Mitra Devkota
From: Tanja Zinßer
To: r-sig-geo@r-project.org
Sent: Wednesday, February 6, 2013 5:53 AM
Subject: [R-si
Dear Gianni,
Also possible is a Method that also works for very large rasters
(calculation done in chunks)
C<-overlay(A, B, fun=function(a,b) return(a==b))
Best
Julian
Am 06.02.2013 13:55, schrieb Wilfried Thuiller:
Dear Gianni,
Something along these lines should directly return what you look for:
C <- A == B
Best
Wilfried
Le 6 févr. 2013 à 13:41, gianni lavaredo a écrit :
> Dear Members list,
>
> I have two raster (A and B) with the same extent, number of rows and
> columns, projection, resolution, a
Dear Members list,
I have two raster (A and B) with the same extent, number of rows and
columns, projection, resolution, and origin. Some pixels of A have the same
values of B.
I wish in a code saving style compare A with B and create a new raster
where the pixel value is 1 (when the statement i
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 06/02/13 12:55, Roger Bivand wrote:
> On Wed, 6 Feb 2013, R.S. Bhalla wrote:
>
>> Dear List, I have run into trouble with the execGRASS command copied below
>> and can't figure
>> out what I'm doing wrong. It runs fine without the size=11 argument
On Wed, 6 Feb 2013, R.S. Bhalla wrote:
Dear List,
I have run into trouble with the execGRASS command copied below and can't
figure out what I'm doing wrong.
It runs fine without the size=11 argument (takes the default value of 3). Can
storage.mode(11)
[1] "double"
is.integer(11)
[1] FALS
Hey everyone,
I'm new in the field of spatial autocorrelation et spatial regression. I'm
searching for a method to run a spatial regression with depend
binary/dichotomous variables.
Does anyone know, if there is a function in R to do so? I only found the
morantest for lm regression. I wanted to a
David, this is a bug that has now been corrected on r-forge.
Thanks for reporting,
On 02/02/2013 04:52 PM, d.g.rossi...@utwente.nl wrote:
I believe some recent change to spplot has removed the possibility of
specifying a printing character other than the default 16 (fill=T) or 1.
The followin
Dear List,
I have run into trouble with the execGRASS command copied below and
can't figure out what I'm doing wrong.
It runs fine without the size=11 argument (takes the default value of
3). Can someone please help.
Thanks in advance,
Ravi
Command:
execGRASS("r.neighbors", parameters=list(inp
Hi,
Someone with a better knowledge of errors like this one might more helpful.
But as a a first start, check the structure of of coords and dat objects (
to see whether there was a mistake while reading coords.txt and dem.txt...
for example headers, or separator issues), then make sure they are o
SpatialPixelsDataFrame() tries to figure out how the points you feed it
with are layed out on a grid structure. Small errors, possibly from
rounding (you're reading from ascii) may cause it to go slightly wrong,
and not give you the grid you'd like.
As you probably know the grid at which "coor
Dear R-Geo users,
I am trying to create a SP objects using SpatialPixelsDataFrame. Next I
can see the code:
> proj <- '+proj=lcc +ellps=WGS84 +lat_0=0 +lat_1=30 +lat_2=60 +lon_0=-4.75'
> coords <- read.table("coords.txt")
> dat<- read.table("dem.txt")
> test <- SpatialPixelsDataFrame(points=c
DO NOT POST HTML!!! All R lists have a plain text only policy, because
plain text uses much less bandwidth than HTML, limiting the carbon
footprint of our list. Servers currently pollute more than most bad
industries! In addition, plain text is much less insecure, because it is
not interpreted
21 matches
Mail list logo
