[R-sig-Geo] help: density map

Marcelino de la Cruz marcelino.delacruz at upm.es
Thu Jul 4 21:26:17 CEST 2013


In general it is acknowledged that it is more important to select a 
bandwith (i.e., a sigma for the Gaussian kernel) than the type of kernel 
(see for example the 1st ed. of the ASDAR book, p. 167).
You can select a sensible bandwith with the functions bw.diggle  in 
spatstat or mse2d in splancs.

By the way, there is a illustrative chapter about kernel functions (in 
spanish) in this book (you can download a free pdf copy)

http://www.bubok.es/libros/225830/Avances-en-el-Analisis-Espacial-de-Datos-Ecologicos-Aspectos-Metodologicos-y-Aplicados


Cheers,
Marcelino

El 04/07/2013 21:02, Ariel Fuentesdi escribió:
> Thank you,
> Why sigma = 10?, what function can I use for choose the best sigma?. And I
> looked the info about density.ppp and it says it use the Gaussian Kernel.
> But, if we look on wikipedia:
> http://en.wikipedia.org/wiki/Kernel_(statistics) , it says, there are more
> kernel function than the Gaussian Kernel, so my question is how do I know
> wether I have to use for example a Gaussian kernel or a Tricube Kernel.
>
> Regards,
> Ariel
>
> 2013/7/4 Juan Tomas Sayago Gomez <juantomas.sayago at gmail.com>
>
>> Hello,
>> you can use the commands in spatstat
>>
>> library(spatstat)
>>
>> summary(swedishpines)
>> plot(density(swedishpines, sigma = 10))
>> then check the help in the spatstat command density.ppp
>>
>> Also check the material in this course http://www.csiro.au/resources/pf16h
>>
>>
>> Juan
>>
>>
>> On Thu, Jul 4, 2013 at 11:32 AM, Ariel Fuentesdi <ariel.fuentesdi at usach.cl
>>> wrote:
>>
>>> Hello,
>>>
>>> I have raingauges as points and I want to make an heatmap with those
>>> points. So, I understand I need to do a kernel density estimation, but,
>>> there are different kernel functions. Thus, my questions is: how do I know
>>> which kernel function I need to apply to my data? Do I need to run a
>>> previous test first?
>>>
>>> I believe I can select the function using grass, but how can I do it with
>>> R? Do I have to use the KernSmooth package? How Can I map that ?
>>>
>>> Regards,
>>> Ariel
>>>
>>>          [[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-sig-Geo mailing list
>>> R-sig-Geo at r-project.org
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>
>>
>>
>>
>> --
>> Juan Tomás Sayago Gómez
>> Graduate Research Assistant
>> West Virginia University - RRI
>> 886 Chestnut Ridge Road, Room 520
>> P.O. Box 6825
>> Morgantown, WV 26506-6825
>>
>
> 	[[alternative HTML version deleted]]
>
>
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>



More information about the R-sig-Geo mailing list