Hello Kad,
The best thing to do is probably to actually *read* those warnings.
It is not that clear what models you tried to fit, but basically I
guess you tested the full range of models available through gstat (see
vgm()). One of them (the bessel model) does not fit using the tested
I you want a more advanced example, you may want to have a look at this US
election map:
http://addictedtor.free.fr/graphiques/RGraphGallery.php?graph=113
(source code here: http://www.ai.rug.nl/~hedderik/R/US2004/ )
Mayeul
_
Dr. Mayeul
Hi Thiago,
My weapon of choice for that would be to use the ggplot2 library.
Here is a quick example. I *hope* it is reproducible (I on't have R
installed on the machine I am posting from):
# here is a sample data frame with locations, a categorical variable
and a continuous variable
dat -
Hi,
I am looking for a function that subsamples spatial data such that points are
beyond a specific distance apart (i.e. all points 1km away from each other; all
points 3km away from each other etc.). Does anybody know if such a function
already exists and if so where I might find it?
Thank
Hi Anna,
Take a look at spDists from the sp-package. After that you probably need
to classify the resulting distances into a few classes, comparable to
the bins for a sample variogram.
cheers,
Paul
On 08/24/2010 03:00 PM, Songhurst, Anna wrote:
Hi,
I am looking for a function that
On 08/23/2010 05:45 PM, Kerry Ritter wrote:
Hi. I am using library gstat in R. My question is why do I get
different SSErr's when I fit the variogram model using fit.variogram
than when I fix the parameters in fit.variogram using the exact same
parameters obtained from the first fit. My
Dear Dr. Hijman,
Thank you for the reply.
Initially I was thinking in using a GIS approach, as a state map would to be
involved in the process.
Please check this link: http://img836.imageshack.us/img836/7021/fig6vdist.png
A work colleague produced this map using ArcGIS, but at a cost of
Dear Mayeul,
Thanks for your tip. I am surely going to analyse the code and try to
reproduce it.
But instead of colouring the municipalities, the only variable I need to work
with would be the indicative circles, something like this
example:
On 08/24/2010 03:58 PM, Thiago Veloso wrote:
Dear Dr. Hijman,
Thank you for the reply.
Initially I was thinking in using a GIS approach, as a state map would to
be involved in the process.
Please check this link:
http://img836.imageshack.us/img836/7021/fig6vdist.png
A work
Dear list members,
I want to plot a neighbours list of class nb. I thought that the plot would
show 8 neighbour lines. But this is not the case for my plot. Did I do
something wrong?
col.knn2nb-knn2nb(knearneigh(coords, k=8, longlat=F))
plot(coords, border=grey)
plot(col.knn2nb, coords,
I've been following this topic with some interest, since this is
something I might like to do fairly often. I'm not that familiar with
ggplot2, but it looks really useful. Is there a way to overplot (or
underplot) a shapefile (e.g. of coastlines) with the approach suggested
by Paul?
Matt
Hi,
I'm new to GIS and have been trying to convert latitude and longitude to/from
state plane coordinates.
I've tried using the project() program from the proj4 library to convert
lat/lng to FIPS 3104 (New York State Long Island).
No matter how I go about this, however, the coordinates come
Hello, Matthew.
Just a superb complement of yours.
I was about to ask the same question to Pierre and Paul, after thanking them
for the useful and functional tip. I managed to follow the ggplot examples, but
a next step would involve plotting my interest points over a shape file (state
On Tue, 24 Aug 2010, Sabira el Messlaki wrote:
Dear list members,
I want to plot a neighbours list of class nb. I thought that the plot
would show 8 neighbour lines. But this is not the case for my plot. Did
I do something wrong?
col.knn2nb-knn2nb(knearneigh(coords, k=8, longlat=F))
On Tue, 24 Aug 2010, Alexandre Villers wrote:
Hello,
Have a look at spTransform (in rgdal package) and the EPSG code of the
desired projection (this is to me the easiest way not to mess with digits and
various copy errors that can be made while writing the projection properties)
at
Hey,
There is an ESRI code (ESRI:102318
http://spatialreference.org/ref/esri/102318/) corresponding to the
requested projection... However, I doubt CRS will take an ESRI code
(Roger ?).
Jonathan, have a look at the rgdal and sp packages help pages for the
How To in CRS()
Best regards
Alex
On Tue, 24 Aug 2010, Alexandre Villers wrote:
Hey,
There is an ESRI code (ESRI:102318
http://spatialreference.org/ref/esri/102318/) corresponding to the
requested projection... However, I doubt CRS will take an ESRI code (Roger
?).
CRS(+init=esri:102318)
but the ESRI version doesn't give
Roger,
This won't really matter since they are very close, but given the data
source and a North America based company, my guess is the underlying
datum is NAD83.
Dan
On 08/24/2010 10:45 AM, Roger Bivand wrote:
On Tue, 24 Aug 2010, Alexandre Villers wrote:
Hey,
There is an ESRI code
I don't know ggplot2 well enough to answer about that, but you can use the
symbols function to add symbols to an existing plot (which could be the map
plotted using maps or maptools/sp etc.). For even more control/options you
could look at my.symbols or subplot in the TeachingDemos package
Hello everyone I am new to the use of R-cran, I have a question in which way
can create arrays of distances (Euclidian and Into network) a network of
roads with space objects.
thanks for your attention
Ricardo Rodríguez
Univalle
[[alternative HTML version deleted]]
hi all am new to the use of R-cran, I have a question that you can carry
objects grassgis calculations or the results of a particular process.
thanks for the help and attention
Ricardo Rodríguez
Univalle
[[alternative HTML version deleted]]
Hello Thiago,
First, I owe you and the list a reproducible example (my first example
does not run properly - my bad):
dat - data.frame(x=runif(50), y=runif(50),
category=sample(letters[1:3], size=50, replace=TRUE),
continuous=runif(50))
# You may want to do some spatial analysis on that
I never thought about duplicates. I have several identical x and y coordinates
for different transactions in an office building. That is why there is no line(
some sort of self neighbour), however not exactly. Thank you very much.
-Oorspronkelijk bericht-
Van: Roger Bivand
Hi:
A couple of things:
(1) It's coord_equal(), not coord-equal() in ggplot2.
(2) For those who want to use Google Maps, I'll just mention that there is
an RgoogleMaps package with a nice vignette on CRAN.
HTH,
Dennis
On Tue, Aug 24, 2010 at 3:17 PM, Pierre Roudier
24 matches
Mail list logo