RELEASE NOTES 

		spatstat 1.13-4

		18 July 2008

	We thank N.A. Laskurain and Marcelino de la Cruz
	for generous contributions.
	
OVERVIEW
	Stepwise model selection can now be performed on point process models. 
	Intersections and unions of polygons now yield a polygon.
	New function kppm() for fitting cluster point process models.
	New dataset added. Improvements to image display.

DEPENDENCE
	Spatstat now depends on the package 'gpclib'. 

NEW DATASET
	urkiola
		Urkiola Woods dataset. 
		A point pattern of two species of trees
		in an irregular polygonal window.
		Kindly provided by N.A. Laskurain and Marcelino de la Cruz.

NEW FUNCTIONS
	
	formula.ppm, terms.ppm
		The generic functions 'formula' and 'terms'
		now have methods for ppm objects (point process models).

	effectfun
		Computes the effect of a single covariate
		in a point process model (ppm object).

	gpc2owin
	owin2pgc
		Conversion between the representations of a polygon
		in the packages spatstat and gpclib.

	kppm
		Fits cluster point process models (Thomas process,
		Matern cluster process), either stationary or with
		inhomogeneous intensity, using an interface similar to ppm.
		There are methods to plot, print, predict and simulate.

IMPROVEMENTS/ENHANCEMENTS

	step, drop1
		It is now possible to perform stepwise model selection
		on ppm objects (fitted point process models), using the
		R commands 'step' and 'drop1'.

	update.ppm
		The update command for ppm objects (point process models)
		now handles formulas involving "." 
		For example update(fit, ~ . + x)

	K-function and its relatives
	(Kest, Kinhom, Kcross, pcf, markcorr, ...)
		These functions now run slightly faster.

	intersect.owin
	union.owin
		The intersection or union of polygons
		now yields a polygon, instead of a binary mask.
		Computations are performed using the package 'gpclib'.

	plot.im
		Explicit argument 'col' controls the colour map. 
		Default colour map changed.

	[.ppp
		The subset index i in X[i]
		can now be a pixel image with logical values.

	lurking
		The first argument can now be a point pattern.

	as.owin
		Now has a method for the class "gpc.poly" of 
		polygons in the package gpclib.

	eval.im
	solutionset
		These functions have a new argument 'envir' 
		specifying the environment in which
		the expression will be evaluated.

BUGS

	[1] "[.ppp"
		If X was an empty point pattern (i.e. containing zero points)
		and W was a window, then X[W] returned X 
		(without resetting the window to W).
		Fixed.

	[2] unmark
		This operator slightly corrupted the internal format
		of ppp objects.
		Fixed.