density /diversity of points

Many thanks for the code!

What are the units of R in marktable? I didn't find the information in 
the documentation.
   

2011/5/17 Marcelino de la Cruz<marcelino.delacruz at upm.es>

   

On 16/05/2011 10:27, Matev?? Pavli?? wrote:

     

Hi all,

I read (just now) about the Simpson idex. This would probably be a 
good thing to try in my case. Do you have any ideas of how to create 
a diversity map using Simpson index?


       

library(vegan)
simpson<- diversity(lansing.tab, "simpson")
lansing.simpson<-  setmarks(lansing,simpson)
lansing.simpson.s<-smooth.ppp(lansing.simpson)
plot(lansing.simpson.s)


  . I have just a few more questions about the code :
     

The line below as I understand sets a table for each point with the 
type of points that are in the near in the radius R=350 units?
mol.tab<- marktable(mol.ppp,R=350)


       

Yes.
To resolve questions like this about the arguments of the functions, 
please read the help pages, e.g.
help(marktable)


  And this line creates a surface(map) of diversity ?
     

lansing.div.s<-smooth.ppp(lansing.div)

       

Yes again.
help(smooth.ppp); help(density.ppp)


     

I think this map that is created with smooth.ppp is to rough giving 
to little detail on the diversity. Would it be possible to use 
kriging to create diversity map?


       

It depends on what do you think is "too rough". You may control 
the"roughnes" of the map changing the "dimyx" argument (that controls 
the final grid of the smoothed surface). You may be also interested 
in setting the argument "sigma" (that controls the bandwith of the smoothing kernel).
Although it is possible to krige the results, I think that in this 
case is preferable to trust the smoothed surface.

Cheers,
Marcelino




  Thanks for the help,
     

m

-----Original Message-----
From: marcelino.delacruz at upm.es [mailto:marcelino.delacruz at upm.es]
Sent: Monday, May 16, 2011 9:59 AM
To: Matev?? Pavli??
Cc: r-sig-geo at r-project.org
Subject: RE: [R-sig-Geo] density /diversity of points

Con fecha 15/5/2011, "Matev?? Pavli&#269;"<matevz.pavlic at gi-zrmk.si>
escribi??:



       

Hi Marcelino,

Was out of the office for a while...
Thanks for the help. I think this could work...but can you tell me 
what this line does?

diversity<- apply(mol.tab,1,function(x) sum(x>0))


         

mol.tab is a table with the number of occurrences of each type 
(columns) in the neighborhood of each point (rows). This line 
computes for each row (i.e. for each point) the number of types whose value is ">0"
(i.e. types that are present in the neighborhood). This is a very 
simple definition of diversity (i.e. "richness"). From that table 
you could also compute Shannon or Simpson diversity indices, if you 
would prefer that.


Marcelino



       

i cant figure out how diversity is calculated here?

Thanks again for the help,

matevz

-----Original Message-----
From: Marcelino de la Cruz [mailto:marcelino.delacruz at upm.es]
Sent: Thursday, May 12, 2011 2:03 PM
To: Matev?...?? Pavli?" Cc: r-sig-geo at r-project.org
Subject: Re: [R-sig-Geo] density /diversity of points

On 12/05/2011 13:12, Matev?...?? Pavli?"  wrote:


         

Hi all,



I have a point data set (SHP) with coordinates and a attribute (i.e.
type of point).

These points are scattered around a fairly big area. What i would 
like to do is to find a sub-area where density of points sombined 
with the diversity of type is the biggest.

Does anyone have any idea iff this is somehowe possible to do in R?
Any idea would be greatly aprpeciated,




           

To your first question:

library(fortunes)
fortune("Yoda")

;-)

More seriously, you could  transform your shp data in a ppp object 
with spatstat. See the vignette in spatstat. Then you can use some 
functions there, for example (with the data set lansing):

library(spatstat)
data(lansing)
plot(lansing)
# get an estimate of point density
lansing.den<- density.ppp(lansing)
plot(lansing.den)

# get an estimate of point diversity (here, for the shake of 
brevity, at the points themselves)
lansing.tab<- marktable(lansing,R=0.05)
diversity<- apply(lansing.tab,1,function(x) sum(x>0)) lansing.div<-
setmarks(lansing,diversity) lansing.div.s<-smooth.ppp(lansing.div)

plot(lansing.div.s)

# select areas with arbitrary high values of  density and diversity plot(
eval.im(lansing.div.s>4.5&   (lansing.den/max(lansing.den))>0.9))


HTH. Cheers,

Marcelino



--

--

	[[alternative HTML version deleted]]