[R] density() with from, to or cut and comparrison of density()

[Rainer M Krug]

Uwe Ligges wrote:
> 
> 
> Rainer M Krug wrote:

SNIP

>> My scenario: I simulated densities of a plants originating from an sseed
>> source at distance zero. Therefore the density of the plants will be
>> highest close to zero. Is there anything I can do to have this pattern?
>> If I use 'from' or 'cut', the resulting densities do not integrate to
>> one which I need as I want to compare different density curves.
> 
> The kernel chosen might be not the ideal one for such a restriction. If
> the density outside the "cut" range is extremely small, you might want
> to do a dirty transformation so that the values sum up to 1 again.
> 

The density on the left of the "cut" range (let's say the left is at 0)
is high - it is as high or even higher then in the left in the estimated
range. But I don't want to include it in any calculations, as I am only
concerned about the decline of the density in the range 0 < x < 125. On
the right "cut" range, the density is nearly zero and in relation to the
estimated range very small.

Which kernel would be a better choice for this kind of scenario?


>> Ny second question is concerning the bandwidth. An I correct in saying
>> that if I want to compare different density estimates that the bandwidth
>> should be the same for all of them?
> 
> Yes.

Thanks.

> 
> Uwe Ligges
> 
> 
>> Thanks in advance for your help,
>>
>> Rainer
>>


-- 
Rainer M. Krug, Dipl. Phys. (Germany), MSc Conservation
Biology (UCT)

Department of Conservation Ecology and Entomology
University of Stellenbosch
Matieland 7602
South Africa

Tel:		+27 - (0)72 808 2975 (w)
Fax:		+27 - (0)21 808 3304
Cell:		+27 - (0)83 9479 042

email:	RKrug at sun.ac.za
      	Rainer at krugs.de