RELEASE NOTES 

		spatstat 1.9-6

	 	27 September 2006

	We thank Colin Beale for suggestions.


OVERVIEW
	Extensions to smoothing code.
	Extension to inhomogeneous K function.	
	Minor improvements.
	
NEW FUNCTIONS

	density.splitppp
		Computes a kernel estimate of intensity 
		for each of the components in a split point pattern. 

	plot.listof
		Plot method for `listof' objects.
		This makes it possible to type, e.g.
			plot(density(split(lansing)))

	summary.profilepl
		method for `summary' for the value returned by profilepl(),
		e.g.         
			s <- data.frame(r=seq(0.05,0.15, by=0.01))
		        ps <- profilepl(s, Strauss, cells)
			summary(ps)

	trim.rectangle
		Simple convenience function for cutting a margin off
		several sides of a rectangle.

IMPROVEMENTS/ENHANCEMENTS

	density.ppp
		Anisotropic smoothing of point patterns is now possible
		using the argument `varcov', which controls the
		variance-covariance matrix of the Gaussian smoothing kernel.
		(Suggested by Colin Beale)

	Kinhom
		If the intensity values lambda are not given,
		they are now estimated using a leave-one-out
		kernel smoother.

	Fest, Gest, Jest
		Computation slightly accelerated.

	eroded.areas
		Computation slightly accelerated.


BUGS

	[1] plot.im
		plot(Z, ribbon=FALSE, axes=FALSE) ignored 'axes'.
		Fixed.

	[2] plot.owin, plot.ppp
		Did not correctly handle the case where
		'main' is an expression.
		Fixed.