[R] "Halfing" Kernels in Kernel Density?

Nils Müller-Scheessel muellerscheessel at rgk.dainst.de
Wed Jan 16 18:11:18 CET 2008

hits=-2.5 tests=BAYES_00,FORGED_RCVD_HELO
X-USF-Spam-Flag: NO


I have got a question regarding the function "density":

Is there a way in R to "half" kernels or to use kernels which are 
heavily skewed to one side?

My problem is as follows: I have a set of archaeological sites which 
are dated by coins (e. g. made in the reign of Hadrianus). However, 
normally, these coins give only the first possible date, but not the 
exact one. So, it is quite possible that the coin was struck in, say, 
115 BC, but was hidden in 160 BC. Such an event is, of course, much 
more likely when the coin was still new. So, the chance that the coin 
got hidden lessened during the course of time.
In order to take this fact in consideration, I would like to use only 
the right half of the, for example, gaussian kernel (or, 
alternatively, a heavily skewed kernel) when plotting the density of 
the coins.

Is there a way to accomplish this in R?

Any suggestions are much appreciated!


More information about the R-help mailing list