Edzer,
This is also my biggest frustration with R at the moment. I can not load
grids that are bigger than 2M pixels and run geostatistics on them (if I run
the same analysis in a GIS software such as SAGA GIS, then I do not get this
problems). I often get messages such as:
Reached total
You can also fit a variogram to the residuals using the gstat package and then
record the nugget and
sill variation (see http://www.gstat.org/manual/node7.html). I do not know how
to test that the
nugget variation is statistically significant from the sill variation. You
could split the
in gstat with skewed distributions
I'm trying to realize e regression kriging with gstat package on my
soil samples data. The response variable (ECe measuere) and covariates
appear positvely skewed.
As Tomislav Hengl suggests in its framework for RK [1], a logistic
transformation is proposed
# Modified by T. Hengl (http://spatial-analyst.net)
# 27.01.2008
library(rgdal)
library(maptools)
library(splancs)
# Import the points and study area:
data.shp - readOGR(C:/, layer=events)
str(data.shp)
poly.shp - readOGR(C:/, layer=hull)
str(poly.shp)
poly -
David,
Thanks for you note. It is a very clear example and I have experienced
this problem many times.
If I can be of any assistance, here are few discussion points that might
put some light on the whole thing.
Gstat does not support any stat models different than linear. This is
mainly
If you import the bands/images as integers [0-255], then you can visualize
them using e.g.:
library(colorspace)
library(rgdal)
vismaps = readGDAL(band_R.tif)
vismaps$red = vismaps$band1
vismaps$green = readGDAL(band_G.tif)$band1
vismaps$blue = readGDAL(band_B.tif)$band1
# Display as a RGB
Dear Roger, Edzer and Rubio,
I would like to join the congratulations (hopefully we will not overload
the mailing list :))) ).
I have been already recommending it to many of my students (I saw the
draft at Edzer's place few months ago).
BTW, the book's website is not available. Is the URL
I completely agree with Thierry.
Take a look at this also:
https://stat.ethz.ch/pipermail/r-sig-geo/2008-February/003176.html
The instructions on how to run RK with binary variables in R you can find in
sec 4.3.3 (Fig. 4.15)
of my lecture notes.
Hengl, T., 2007. A Practical Guide to
The two components of the regression-kriging model are not independent, hence
you are doing a wrong
thing if you are just summing them. You should use instead the universal
kriging variance that is
derived in gstat. The complete derivation of the Universal kriging variance is
available in
(estimation
error);
But I guess that a method to directly back-transform the UK variance from logit
scale to 0-1 scale
does not exist.
Tom Hengl
-Original Message-
From: Edzer Pebesma [mailto:[EMAIL PROTECTED]
Sent: woensdag 9 april 2008 14:48
To: Tomislav Hengl
Cc: r-sig-geo
Auxiliary variables that are used to explain the trend-part of variation need
to be available also
at all new prediction locations.
see ?krige:
newdata - data frame or Spatial object with prediction/simulation locations;
should contain
attribute columns with the independent variables (if
)/cellsize)
yN - round((yUR-yLL)/cellsize)
data.grid - expand.grid(x=seq(xLL,xUR,xN), y=seq(yLL,yUR,yN))
names(data.grid) = c(X,Y)
gridded(data.grid) - ~X+Y
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of
Tomislav Hengl
Sent: maandag 14 april 2008 10:37
Dear list,
I know that much has already been said about the memory limit problems. If
there is any progress
about this problem, we would be interested to hear.
In our project, we are importing 24 maps/bands, each consists of 1,450,000
pixels. We further would
like to glue all maps into a
: dinsdag 22 april 2008 16:14
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch; 'Michalis Vardakis'
Subject: Re: [R-sig-Geo] Memory limit problems in R / import of maps
Hi Tom,
Tomislav Hengl wrote:
Should we simply give up on running spatial analysis using large grids (10
million grids) in R
Beaudette [mailto:[EMAIL PROTECTED]
Sent: dinsdag 22 april 2008 17:22
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch; Michalis Vardakis
Subject: Re: [R-sig-Geo] Memory limit problems in R / import of maps
On Tue, Apr 22, 2008 at 6:49 AM, Tomislav Hengl [EMAIL PROTECTED] wrote:
Dear list,
I know
to work on the R code).
Tom
-Original Message-
From: milton ruser [mailto:[EMAIL PROTECTED]
Sent: dinsdag 22 april 2008 18:21
To: Tomislav Hengl
Cc: Dylan Beaudette; r-sig-geo@stat.math.ethz.ch; Michalis Vardakis
Subject: Re: [R-sig-Geo] Memory limit problems in R / import of maps
Hi
The Computational Geo-Ecology (CGE) research group of the University of
Amsterdam is looking for two
enthusiastic Scientific Programmers who will develop virtual laboratories that
support data analyses
and modelling. This concerns data exploration through
***
From 25-30 August 2008 the research school ICG organizes a five-day course
targeted at PhD students,
academic staff and project teams working with spatio-temporal data. The course
aims at providing a
balanced combination of theoretical and hands-on-software
If it is of any help, you might want to take a look at this guide:
Creating Maps for Publication using R Graphics
http://www.nceas.ucsb.edu/scicomp/GISSeminar/UseCases/MapProdWithRGraphics/OneMapProdWithRGraphics.h
tml
The second example demonstrates how to visualize a remote sensing band using
FYI
I have just finishing preparing an R script that accompanies my paper in
Computers in Geosciences on
finding the grid cell size for various applications.
The script and the data sets can be obtained from:
http://spatial-analyst.net/pixel.php
This gives a step-by-step guide to estimate
For European projects, we commonly use the official European Terrestrial
Reference System
(www.euref.eu). You can convert your data from the longlat system to the ETRS
using:
library(maptools)
proj4string(pointmap) - CRS(+proj=longlat +datum=WGS84)
pointmap.etrs - spTransform(pointmap,
Dear Dave,
I separate fitting of the deterministic (trend) and residual part of the
universal kriging model all
the time. Adding OK of residuals to the trend is fine, as long as the
regression model is estimated
using GLS (but many do it even if they use only OLS; the difference is often
I forgot to mention - the book's website is already on-line:
http://www.bias-project.org.uk/asdar-book/
The datasets and R scripts are still missing, but I imagine it should all soon
be there? (for me
personally, R scripts/datasets/examaples are equally important as the book
itself)
cheers,
Message-
From: Gomez Rubio, Virgilio [mailto:[EMAIL PROTECTED]
Sent: donderdag 10 juli 2008 16:40
To: Tomislav Hengl; [EMAIL PROTECTED]
Cc: r-sig-geo@stat.math.ethz.ch
Subject: RE: [R-sig-Geo] How to get Applied Spatial Data Analysis withRbefore
9th of August
Hi,
It is good to see that the book
Yes, buffer distances can be smoothly calculated using SAGA (I do not if it
will work for your HUGE
image, but you can try). Here is an example:
.
# Import the contour data (from the 1:5k and 1:50k topo maps):
library(maptools)
getinfo.shape(contours.shp)
contours.50 -
You can also try RSAGA. It is very fast and handy to up-scale large grids (it
is perfect to automate
processing), and you only need to define the new grid cell size and the
up/down-scaling method:
library(RSAGA)
# Download SAGA GIS from www.saga-gis.org and unzip the binaries to:
Dear Radim, Edzer,
I was thinking about the same problem few years ago (I assume that you work
with auxiliary maps and
not only coordinates).
I think (have a feeling) that local and global Universal kriging should be
treated as two things
(especially if you put a very narrow search radius).
try this:
***
library(akima)
library(spatstat)
library(maptools)
ew - as.vector(c(1394180, 1398190, 1399690, 1395780, 1397820, 1395290,
1394995, 1399475, 1410825))
ns - as.vector(c(6200295, 6201020, 6200420, 6205105, 6206590, 6208305,
6210370, 6213715, 6214870))
cures -
Kernel density can be derived in spatstat package, but also in adehabitat and
many other packages.
You need to loop your operations (output as a list). Note that I use the same
file names to save
space - you only need the output of course!
***
library(maptools)
library(rgdal)
Dear Roger, Edzer and Virgilio,
Somehow, I was not lucky enough to receive a copy of your ASDAR book (I only
had a chance to browse
it yesterday at Edzer's place). First, I followed the R course Edzer gave at
GEOENV2008, and our
copy did not arrive on time (the organizers made a wrong shipment
-analyst.net
-Original Message-
From: Virgilio Gomez Rubio [mailto:[EMAIL PROTECTED]
Sent: Wednesday, October 22, 2008 12:34 PM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Problems with obtaining a copy of your ASDAR-book
Dear Tom,
Many thanks for your
Yes, you can export sgrd to geoTIFFs via the io_grid_gdal module:
library(RSAGA)
rsaga.get.modules(io_grid_gdal)
$io_grid_gdal
code name interactive
10Import Raster via GDAL FALSE
21Export Raster via GDAL FALSE
32
If the problem is memory handling in R (still a big headache to do much of GIS
analysis in R), you
should instead consider reading pieces of grids that you need e.g.:
info - GDALinfo(dem50m.asc)
info
gridmaps01 = readGDAL(dem50m.asc, region.dim =
round(c(info[[rows]]/2,info[[columns]])))
to get statistics of grids over polygons, you might also consider using the
Grid Statistics for
Polygons method in SAGA GIS:
rsaga.get.modules(shapes_grid)
$shapes_grid
code name interactive
1 0Add Grid Values to Points FALSE
2 1 Get
Dear Marta,
I am not sure what is the projection system you use for the Algarve case study?
If it is different
from CRS(+proj=longlat +datum=WGS84), then you do need to first reproject
your maps.
You should really look at the examples from the book by Bivand et al.:
Alessandro,
You need to separate the put the list of input grids as a character vector e.g.:
GRIDS=c(DSM1.sgrd,DSM2.sgrd,DSM3.sgrd)
See also:
http://spatial-analyst.net/wiki/index.php?title=Software#SAGA_GIS
Maybe you should also study this posting guide:
Hi Alessandro,
Please do submit such and similar bugs to Olaf Conrad (the main developer of
SAGA). For your
information, the C++ code of SAGA (SAGA 2.0.2 API - Python Interface) is
available from
http://sourceforge.net/projects/saga-gis/
which means that you could obtain the code and fix the
Dear Holger,
This is a small script that uses [colorspace] library to run calculations with
colors (some methods
are limited to 256 colors only):
http://spatial-analyst.net/scripts/whitening.R
If you assign R,G,B values spatially and then interpolate them (they have to be
in the 0-1 range),
You could also loop e.g. (this way you have a bit more control):
long - c(-81.66,-82.66)
lat - c(36.21,37.21)
df - data.frame(lat,long)
df$srtm - rep(NA, length(df$lat))
for(i in 1:length(df$srtm)){
+ df$srtm[i] - GNsrtm3(df$lat[i],df$long[i])$srtm3
+ }
df
lat long srtm
1 36.21
Dear Pritam,
Please forward us the R code, and, if possible, upload your data to some ftp
server.
The coordinate system string of the maps imported to R can be set using:
data(meuse)
coordinates(meuse) - ~x+y
proj4string(meuse) - CRS(+init=epsg:28992)
A list of epsg code is available via
Apologies for cross postings!
FIRST CALL FOR PAPERS
Geomorphometry 2009
29 August - 2 September 2009
Zurich, Switzerland
http://2009.GEOMORPHOMETRY.ORG
e-mail: [EMAIL PROTECTED]
PROGRAM CHAIRS
Ross Purves University of Zurich
Stephan Gruber University of Zurich
Tomislav Hengl
A quick way to rescale a set of maps is to use the grid tools of SAGA GIS (see
rsaga.get.usage(grid_tools, 0)). First, download all ASCII file maps to a
working directory, then
obtain the list of maps in the folder, and then automate:
ascmaps - list.files(getwd(), pattern=\\.asc$, full=F)
Dear Henk,
My experiences is that the easiest way to read KML files and convert them to
sp-type of objects is
to use the XML package.
Here is a small example:
library(XML)
meuse_lead.kml - xmlTreeParse(meuse_lead.kml)
lengthp - length(meuse_lead.kml$doc[[1]][[1]][[1]])-1
lead_sp -
Dear Alok,
Maybe you should consider converting the im data to sp class. Then, you have as
many grids in a
single SpatialGridDataFrame e.g.:
library(maptools)
library(rgdal)
grids - as(bei.extra[[1]], SpatialGridDataFrame)
names(grids)[1] - elev
grids$grad - as(bei.extra[[2]],
-Original Message-
From: b.rowling...@googlemail.com [mailto:b.rowling...@googlemail.com] On
Behalf Of Barry
Rowlingson
Sent: Tuesday, January 20, 2009 10:17 AM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Spatial data analysis in MATLAB / Comparison
Hi Barry, Edzer,
Just to let you know that geonames.org server changed from ws.geonames.org to
ws5.geonames.org
See the info at: http://www.geonames.org/export/
Maybe you should add to your R package geonames an option to specify the
server name manually,
e.g.:
setGNserver -
Hi all,
I have just finished preparing a small repository of publicly available Global
maps on environment.
For more info see:
http://spatial-analyst.net/wiki/index.php?title=Global_datasets
The maps can be downloaded directly from:
http://spatial-analyst.net/worldmaps/
I would be
Dear Herry,
If I understand what you problem, one solution is to use R+SAGA. You should
first convert the
polygon map to the same grid, and then you can load it to R and do any type of
aggregation:
library(maptools)
library(rgdal)
library(RSAGA)
# load the gridded map:
rastermap -
...@stat.math.ethz.ch] On Behalf
Of Tomislav Hengl
Sent: Wednesday, January 28, 2009 9:56 AM
To: r-sig-geo@stat.math.ethz.ch
Cc: alexander.h...@csiro.au
Subject: Re: [R-sig-Geo] calculate raster values based on vector regions
Dear Herry,
If I understand what you problem, one solution is to use R
Dear Yong Li,
I hope you will not mind me joining this interesting discussion.
If there is no evident spatial auto-correlation structure (pure nugget effect),
IDW/OK are as good
as randomly drawing a value from the global (normal) distribution. You can even
test this using
cross-validation!
Dear Rainer,
This is of course possible in R, and can be done in several ways:
1) for example, you can derive the average value using the rowSums function:
maps$Nsum - rowSums(m...@data, na.rm=T, dims=1)
maps$avg - maps$Nsum/(length(names(meuse.g...@data))-1)
You could also loop the sd,
Jonathan,
I think this is an excellent idea.
Ideally we should have a system that automatically scans e-mails, distinguishes
between a question,
answer/suggestion, and R code, and then generates a wiki article (such articles
could then be
classified into some hierachical structure using
Hi Barry,
SAGA has few simple but useful modules for extraction of features from grids.
See:
library(RSAGA)
rsaga.get.modules(grid_discretisation)
$grid_discretisation
code name interactive
10 Supervised Classification FALSE
21 Cluster Analysis for
Dear R-sig-geo,
There are still some open places for our GEOSTAT 2009 summer school (the
registration closes on 15h
of March). We will finally be three lecturers: (1) Roger Bivand; (2) Olaf
Conrad (the main creator
of SAGA) and (3) Tomislav Hengl. The fifth day of the summer school, we
Dear Else,
I've read your message twice, but I am still not sure what exactly you want to
run: (1) which
prediction technique, (2) for which purpose (validation only?).
Using the krige.cv method, you can also run the cross validation for inverse
distance interpolation
(note: you only need to
Dear list,
I am trying to test some DEM analysis functions in GRASS GIS (under MS Windows
machines). This is my
first contact with GRASS scripting (I have successfully installed GRASS 6.4 for
windows from
http://grass.itc.it/grass63/binary/mswindows/native/).
For example, with SAGA GIS, I can
, March 25, 2009 10:46 AM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch; carlos.grohm...@gmail.com
Subject: Re: [R-sig-Geo] Running analysis on DEMs using GRASS (from R?)
Tomislav,
This would be part of an R script using grass:
brachspread - function(itmax=1,a=-1)
{
#Do the next once per
Message-
From: Roger Bivand [mailto:roger.biv...@nhh.no]
Sent: Wednesday, March 25, 2009 1:27 PM
To: Agustin Lobo
Cc: Tomislav Hengl; carlos.grohm...@gmail.com; r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Running analysis on DEMs using GRASS (from R?)
On Wed, 25 Mar 2009, Agustin
Hi Greg,
Yes, there are many possibilities for downscaling grids in R, so you are at the
right place. :)
1. If you only wish to downscale climatic grids (e.g. using splines), then
probably the most
efficient (fastest; can handle large grids) way is to use the downscaling
method in SAGA:
-Original Message-
From: r-sig-geo-boun...@stat.math.ethz.ch
[mailto:r-sig-geo-boun...@stat.math.ethz.ch] On Behalf
Of gianni lavaredo
Sent: Tuesday, April 07, 2009 4:44 PM
To: r-sig-geo@stat.math.ethz.ch
Subject: [R-sig-Geo] problem to module MERGE GRID in RSAGA
Dear R Users
Hi all,
Thanks to Markus Loecher and colleagues we can now easily obtain background
maps from Google Earth
and use it for plotting/interpretation of spatial data.
This runs pretty smoothly (e.g. a map of the Netherlands):
library(RgoogleMaps)
# obtain the API key and save into the home
Title: Automated analysis of elevation data in R+SAGA/GRASS
Venue: University of Zurich, Irchel Campus, 29 August 30 August 2009
Workshop moderators: Tomislav Hengl / Carlos H. Grohmann
Registration fees: 150 CHF (PhD students)
For more info: http://2009.geomorphometry.org/
Summary
FYI: The official report from the GEOSTAT 2009 Summer School for PhD students:
http://spatial-accuracy.org/FromGEOSTAT2009
TITLE: Spatio-temporal data analysis with R+SAGA+Google Earth
LOCATION: Mediterranean Institute for Life Sciences, Split, Croatia
PERIOD: 3-10 May 2009
LECTURERS: R.
]
Sent: Thursday, June 11, 2009 11:02 AM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Problems with loading the mgcv package
In general, before you post questions of this kind to the list, it makes
sense to see if the packages load on your platform in a clean
Tom Hengl
http://spatial-analyst.net
-Original Message-
From: Thomas Adams [mailto:thomas.ad...@noaa.gov]
Sent: Friday, July 10, 2009 4:20 PM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Slope and Aspect calculations in R
Tom,
Thanks! I have seen
Hi Agustin,
You should really work with some referent point for which you know both local
and geographic (WGS84)
coordinates.
Here is an illustration:
http://spatial-analyst.net/wiki/index.php?title=MGI_/_Balkans_coordinate_systems#Validation_of_CRS_p
arameters
Note that I use Google Earth
: Tuesday, July 14, 2009 4:53 PM
To: Tomislav Hengl
Cc: alexandre villers; r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] getting subfolders in a directory on FTP (MODIS)
On Tue, Jul 14, 2009 at 2:12 PM, Tomislav
Henglhe...@spatial-analyst.net wrote:
To tell you honestly, I could
-Original Message-
From: r-sig-geo-boun...@stat.math.ethz.ch
[mailto:r-sig-geo-boun...@stat.math.ethz.ch] On Behalf
Of Alexandre VILLERS
Sent: Wednesday, July 15, 2009 10:03 AM
To: Aide R SIG GEO
Subject: [R-sig-Geo] still struggling with FTP download on MODIS
Dear all,
I'm
Dear Emmanuel,
I have the same problem. I either can not run processing with large data set in
R or I can not even
load such data to R. Then, if I want to do any geostatistics, it takes forever.
R (gstat/geoR) is
simply not that efficient with large spatial data as e.g. GIS software.
What you
Hi Laura,
Sorry for a bit delay in reply. I hope you did not give up on geoR. The truth
is: transforming
result of prediction using geoR to e.g. sp classes is not trivial (it looks
like as creators of geoR
did not really consider this option). Once you get it into sp class, then it is
easy to
SAGA has a module to convert from shapes to grids (GRASS can do the same).
See:
library(RSAGA)
rsaga.get.usage(lib=grid_gridding, module=0)
SAGA CMD 2.0.3
library path: C:/Progra~1/saga_vc/modules
library name: grid_gridding
module name : Shapes to Grid
Usage: 0 [-GRID str] -INPUT str
Dear R-sig-geo,
I have just finished with updating the second edition of my lecture notes
that I have used over years for the GEOSTAT summer schools for PhD
students.
A copy of the working draft can be obtained here:
http://spatial-analyst.net/book/
I would like to invite you to read
Here is another example with overlay from the Reimann et al. (2008) book
(this uses a raster map as background - this if often more impressive):
http://www.statistik.tuwien.ac.at/StatDA/R-scripts/page151.html
The code is rather long, but it will give you good ideas what you have to do.
HTH
T.
Dear Alexandre,
I think that this should be doable.
Once you run the ENFA to estimate the habitat model, copy the values of new
rasters (future habitat)
and replace the original values in the object. Then simply re-run the
predict.enfa method e.g.:
# prepare the dataset for ENFA:
beidata -
this helps,
Mathieu.
Tomislav Hengl a écrit :
Dear Alexandre,
I think that this should be doable.
Once you run the ENFA to estimate the habitat model, copy the values of
new rasters (future habitat)
and replace the original values in the object. Then simply re-run the
predict.enfa method e.g
Hi Abe,
Did you try using the ETOPO1? This (0.1 degree) DEM can be well loaded to R:
URL - http://spatial-analyst.net/worldmaps/;
# list of maps:
map.list - c(globedem, chlo08, dcoast)
# download the zipped maps one by one:
for(i in 1:length(map.list)) {
download.file(paste(URL,
-Original Message-
From: Aben Woha [mailto:abenw...@gmail.com]
Sent: Tuesday, September 15, 2009 7:52 AM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] DEMs
Hi Tomislav,
Thanks for the detailed response. I tried to use the globedem.zip from your
Jonas,
SAGA (GUI) can read and write shape files. This is the native vector format it
uses. Maybe this is
confusing, but RSAGA is NOT a GIS package under R, but only a link to send
things from R to SAGA and
back.
Here are some examples:
Hi Harry,
The key issue is that you need to select the (Euclidean) coordinate system. For
example, to
represent the whole world, there are many possiblities
(http://www.radicalcartography.net/?projectionref). If you are working with the
great circle
distance, make sure that the geographic
-Original Message-
From: Roger Bivand [mailto:roger.biv...@nhh.no]
Sent: Friday, October 16, 2009 11:06 AM
To: Tomislav Hengl
Cc: r-sig-geo@stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Including SAGA grid as GDAL-supported format
On Fri, 16 Oct 2009, Tomislav Hengl wrote
Hakim,
There are plenty of (SDM) algorithms that generate species
distribution maps given occurrence (or density) records and
environmental maps. For a systematic review, see e.g. Tsoar et al.
2007 [http://dx.doi.org/10./j.1472-4642.2007.00346.x].
In R, you can try using the
Marcelo,
Surprisingly, I could not find any function in the spatstat package (or splancs
package) that
specifically derives cross-correlations between multiple point processes:
data(lansing)
plot(split(lansing)) # distribution of occurrence records for five+1 species;
Roger Bivand wrote:
On Tue, 10 Nov 2009, ageel bushara wrote:
Dear list members,
I'm doing spatial interpolation of soil moisture using kriging with
external drift following the example given by Tom Hengl
(http://spatial-analyst.net/wiki/index.php?title=Regression-kriging_guide).
When I
Dear Edzer, Jon, Stephanie, Olivier et al.,
First, let me congratulate you on your new package intamap /
intamapInteractive that I feel will significantly advance practice of
geostatistical mapping. Really a great contribution (I knew about SSA
already in 2000, but was always lacking an
Hi Nick,
Very interesting problem. At first thought, I imagined that you just want to
simulate noise ;)
In the geoR package (http://leg.ufpr.br/geoR/) there is a function to simulate
Gaussian Random
Fields (uses actually RandomFields package) using various models e.g.:
library(geoR)
Hi Paul, Edzer,
I understand why the singular matrix problem happens and I know that there is
not really a
mathematical solution around it:
x - matrix(runif(100), nrow=10)
x.i - solve(x)
str(x.i)
num [1:10, 1:10] 0.8191 -1.0293 0.0826 1.068 -0.2106 ...
# add a 'singular' column
x[,1:10] -
Hi Ultrich,
Facundo Muñoz apparently made a GRASS function to derive distances along
streams:
https://stat.ethz.ch/pipermail/r-sig-geo/2009-November/006851.html
17 observations only? That is really tight for any geostatistical analysis (on
top, you want to do 3
dimensions!). I would instead
To: Tomislav Hengl
Cc: 'r-sig-geo'
Subject: Re: [R-sig-Geo] 'LDLfactor' error in 'krige' function
Tom, you can already do this:
library(gstat)
data(meuse)
coordinates(meuse)=~x+y
data(meuse.grid)
gridded(meuse.grid)=~x+y
meuse=meuse[c(1,1:155),] # replicate first observation
pr1 = predict(zinc
Hi Cara,
There is a very efficient function in SAGA called Close Gaps that does
exactly that. What makes it
especially efficient is that it allows you to set a mask map. See:
rsaga.get.usage(grid_tools, 7)
SAGA CMD 2.0.3
library path: C:/Progra~1/saga_vc/modules
library name: grid_tools
Hi all,
FYI: I've run a small comparison between ASTER DEM and LIDAR DEMs (say 'true'
topography):
http://geomorphometry.org/content/gdem-quick-assessment
The 4 case studies can be downloaded from here:
http://geomorphometry.org/system/files/GDEM_assessment.zip
I'm not too happy with what
Dear Pete, Edzer,
If this is of any help, few years ago we have played with using auxiliary maps
to interpolate
categorical variables (multi-indicators?). The results are reported in:
Hengl T., Toomanian N., Reuter H.I., Malakouti M.J. 2007. Methods to
interpolate soil categorical
variables
David Rossiter runs this excellent distance education course:
http://www.itc.nl/Pub/Study/Courses/C10-AES-DE-02
*Next one starts in 6 weeks!
We will run a 5-day trainig course in June (which is more affordable,
but shorter):
27 June - 4 July 2010, GEOSTAT 2010 Summer School for PhD
/experiences are welcome (before we
start installing and testing the functionality).
Thanx!
Tomislav Hengl and Lourens Veen
___
R-sig-Geo mailing list
R-sig-Geo@stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Angera - Italy
email: stef...@casalegno.net
web: http://www.casalegno.net/stefano
On Dec 13, 2009, at 6:29 PM, Tomislav Hengl wrote:
Dear R-sig-geo,
As a part of our project (EcoGRID.nl) we have prepared some 60
thematic grids that we use
(uni-muenster.de), Gerard Heuvelink
(wur.nl), Olaf Conrad (geowiss.uni-hamburg.de), Tomislav Hengl (uva.nl)
and Victor Olaya Ferrero (unex.es)
-
DEADLINE TO REGISTER:
15th February 2010; online at:
http://geostat2010.info
Tobin Cara wrote:
Hello,
I have recently read an interesting article about integrating Limited Area
Models (LAMs) into kriging with external drift for temperature (Libert� et al.
link below).
www.wmo.int/pages/prog/www/IMOP/.../IOM.../P2(05)_Perini_Italy.doc
This URL is incomplete. Please
Dear Laure,
If this is still actual, here are few tips on how to aggregate continuous
information (rasters) given some polygon maps.
-
# obtain a polygon map e.g. World countries:
load(url(http://spatial-analyst.net/DATA/WorldPolyCountries.Rdata;))
/
-Original Message-
From: r-sig-geo-boun...@stat.math.ethz.ch [mailto:r-sig-geo-
boun...@stat.math.ethz.ch] On Behalf Of Tomislav Hengl
Sent: Monday, December 28, 2009 9:50 AM
To: r-sig-geo
Cc: Victor Olaya; Olaf Conrad
Subject: [R-sig-Geo] GEOSTAT 2010 Summer School, 27 June - 4 July
Here are several data sets that I try to maintain and improve always:
DEMs for testing geomorphometry algorithms:
http://geomorphometry.org/content/data-sets
The Baranja hill data set includes many rasters of different type:
http://geomorphometry.org/content/baranja-hill
We used in in our
Dear Jonathan,
Interesting topic. Please send us also your code example and if possible part
of your dataset (I assume it is a SpatialGridDataFrame?).
Operations on lists can be speed up by using lapply and by implementing your
own functions, consider also running Rprof to check which
1 - 100 of 118 matches
Mail list logo