[R-sig-Geo] "one-dimensional" spatial point process

[Adrian.Baddeley at csiro.au]

Matilde Karakachoff <karaka at ifc.cnr.it> writes: 

>  I'm looking for some simple techniques of exploratory data analysis for a
> "one-dimensional" spatial point process. For example I would test the CSR of
> some (few) points in a line of length 5000 against the hp. that they cluster
> in one or two places of the line.

> Do you think that there exists an r-package or function to solve this kind
> of problem? I really like spatstat, but I don't think that I could use it in
> a "one-dimensional" case.

The spatstat package has facilities for exploratory analysis of point patterns on a linear network
(see help for 'lpp'). A special case is a point pattern on a single line. See help(spatstat) under 'Linear Networks' for a list of 
functions available. 

Of course if you are willing to assume the points are Poisson, and you want to distinguish CSR from inhomogeneous Poisson, you can do this with standard methods for random variables - e.g. the Kolmogorov-Smirnov test of goodness-of-fit to the uniform distribution; density estimation in 1D; etc

Adrian Baddeley