RELEASE NOTES

spatstat 1.44-0

22 December 2015

We thank Jonas Geldmann, Andrew Hardegen, Kassel Hingee, Tom Lawrence, Robin Milne, Gopalan Nair, Suman Rakshit, Peijian Shi and Rasmus Waagepetersen for contributions.

OVERVIEW

NEW FUNCTIONS

SIGNIFICANT USER-VISIBLE CHANGES

BUG FIXES

  1. runifpointx, rpoisppx Crashed if nsim > 1. Fixed.

  2. triangulate.owin Results were incorrect in some special cases. Fixed.

  3. quadrat.test, clarkevans.test In rare cases, the computed Monte Carlo p-value could have been greater than 1. This could have occurred only when nsim was an even number and when the correct p-value was equal to 1. Fixed.

  4. linearmarkequal Result was a data frame instead of an ‘fv’ object. Fixed.

  5. point-in-polygon test The function inside.owin could take a very long time to check whether points are inside a polygonal window, if the coordinates were very large numbers. This was due to numerical overflow. (Fixed??)

  6. as.fv.kppm Crashed if the model was not fitted by minimum contrast. Fixed.

  7. plot.fv Crashed in some obscure cases. Fixed.

  8. collapse.fv Did not allow ‘same=NULL’. Fixed.

  9. dclf.progress, mad.progress, dg.progress, dclf.sigtrace, mad.sigtrace, dg.sigtrace The results could not be re-plotted using a plot formula, because the internal data were slightly corrupted. Fixed.

  10. Kest.fft Result was incorrectly normalised. Fixed.

  11. crosspairs If X and Y were identical point patterns, the result was not necessarily symmetric (on some machines) due to numerical artifacts. Fixed.

  12. plot.fv Lines were not correctly clipped to the plot region when ‘ylim’ was given. Fixed.

  13. pool.envelope The ‘scale’ argument was not handled correctly. Fixed.

Release notes are available in raw text format here.