RELEASE NOTES 

		spatstat 1.12-8

		26 February 2008

	We thank Dominic Schuhmacher for detecting several bugs.


OVERVIEW
	Substantial revision of area-interaction models
	including bug fix in simulation code.

IMPROVEMENTS

	suffstat
		The second argument X now has a sensible default.

	AreaInter

		The area-interaction model has been reformulated using
		a new set of parameters.

		This should make the fitted models easier to understand, 
		and improve numerical stability of the fitting algorithm.

		The change of parameters affects both 

			- the fitting of area-interaction models 
				using ppm and AreaInter

			- the simulation of area-interaction models
				using rmh 

		To interpret the new parameters, see help(AreaInter).


COMPATIBILITY

	AreaInter

		Point process models (objects of class "ppm") involving the
		area interaction, such as
			ppm(swedishpines, ~1, AreaInter(2))
		that were fitted in spatstat 1.12-7, are now in an
		incorrect format. These objects should be discarded
		and the models re-fitted in spatstat 1.12-8.


BUGS

	[1] rmh
		The code for simulating the area-interaction model
		contained a bug in the evaluation of the conditional
		intensity. Hence, simulated realisations of area-interaction
		models were completely incorrect.
		Fixed.

	[2] suffstat, 
            ppm
		[Under R 2.6.1 and 2.6.2 only]
		For point process models fitted using the interactions
		OrdThresh() or AreaInter() without a spatial trend,
		the output of suffstat() and ppm(...., method="ho")
		was slightly incorrect (slightly underestimating the
		intensity parameter).
		Fixed.