RELEASE NOTES 

		spatstat 1.12-9

		17 March 2008

	We thank Devin Johnson, Hao Wang and Jeff Picka for contributions.

OVERVIEW
	Extensions and changes to model-fitting capabilities.
	Improvements to robustness of diagnostic code.

NEW FUNCTIONS

	model.matrix.ppm
		Method for model.matrix() for ppm objects.
		Extracts the design matrix from a fitted
		spatial point process model.

	model.images
		Analogous to model.matrix.ppm except that
		it computes pixel images of the covariates,
		instead of computing the covariate values 
		at certain points only.
		Suggested by Devin Johnson.

	SatPiece
		creates a piecewise-constant saturated pairwise interaction
		for use in fitting point process models using ppm(). 
		This interaction is a generalisation of Geyer() 
		to the case of several interaction distances.
		Studied by Hao Wang and Jeff Picka in collaboration with
		the authors.

IMPROVEMENTS

	ppm
		The default value of `rbord', the border correction radius,
		has been changed. Previously the default was rbord=0.
		The default is now rbord=reach(interaction), 
		the interaction distance of the interpoint interaction in 
		the model. This ensures that the maximum pseudolikelihood
		estimating equations are unbiased in the default case.

	lurking
		Improved error handling
		
	diagnose.ppm
		Improved argument handling


BUGS

	[1] Geyer
		The interaction range of the Geyer saturation process
		was incorrectly computed as reach(Geyer(r,sat)) = r.
		The correct value is 2*r.
		Fixed.

	[2] lurking
		Did not detect incorrect values of argument 'type',
		and issued an obscure error message.
		Fixed.