Spatstat

analysing spatial point patterns

• Home
• News
• Download
• Docs
• Book
• Courses
• Help
• FAQ
• About

RELEASE NOTES

spatstat 1.45-0

10 March 2016

We thank Monsuru Adepeju, Mario D’Antuono, Markus Herrmann, Paul Hewson, Kassel Hingee, Greg McSwiggan, Suman Rakshit and Sven Wagner for contributions.

OVERVIEW

• Important bug fix in leverage/influence diagnostics for Gibbs models.

• Numerous bug fixes in code for replicated point patterns.

• Surgery on linear networks.

• Tessellations on a linear network.

• Laslett’s Transform.

• Colour maps for point patterns with continuous marks are easier to define.

• Pair correlation function estimates can be pooled.

• Stipulate a particular version of a package.

• Fixed namespace problems arising when spatstat is not loaded.

• Bug fixes and performance improvements.

• spatstat now contains 100,000 lines of R code.

• Version Nickname: ‘One Lakh’

NEW FUNCTIONS

• laslett Laslett’s Transform. [Thanks to Kassel Hingee]

• lintess Tessellation on a linear network.

• divide.linnet Divide a linear network into pieces demarcated by a point pattern.

• insertVertices Insert new vertices in a linear network.

• thinNetwork Remove vertices and/or segments from a linear network etc

• connected.linnet Find connected components of a linear network.

• nvertices, nvertices.linnet, nvertices.owin Count the number of vertices in a linear network or vertices of the boundary of a window.

• as.data.frame.linim, as.data.frame.linfun Extract a data frame of spatial locations and function values from an object of class ‘linim’ or ‘linfun’.

• as.linfun, as.linfun.linim, as.linfun.lintess Convert other kinds of data to a ‘linfun’ object.

• requireversion Require a particular version of a package (for use in stand-alone R scripts).

SIGNIFICANT USER-VISIBLE CHANGES

• [.linnet, [.lpp New argument ‘snip’ determines what to do with segments of the network that cross the boundary of the window. Default behaviour has changed.

• pcfinhom Default behaviour is changed when ‘lambda’ is a fitted model. The default is now to re-fit the model to the data before computing pcf. New arguments ‘update’ and ‘leaveoneout’ control this.

• envelope methods New argument ‘funYargs’ contains arguments to the summary function when applied to the data pattern only.

• plot.ppp, plot.lpp For a point pattern with continuous marks (‘real numbers’) the colour arguments ‘cols’, ‘fg’, ‘bg’ can now be vectors of colour values, and will be used to determine the default colour map for the marks.

• symbolmap Now accepts a vector of colour values for the arguments ‘col’, ‘cols’, ‘fg’, ‘bg’ if argument ‘range’ is given.

• closepairs.ppp, closepairs.pp3 New arguments ‘distinct’ and ‘neat’ allow more options.

• closepairs.ppp, closepairs.pp3 Argument ‘ordered’ has been replaced by ‘twice’ (but ‘ordered’ is still accepted, with a warning).

• closepairs.ppp, closepairs.pp3 Performance improved (computation time and memory requirements reduced.) This should improve the performance of many functions in spatstat.

• Geyer The saturation parameter ‘sat’ can now be less than 1.

• lpp, as.lpp These functions now handle the case where ‘seg’ and ‘tp’ are given but ‘x’ and ‘y’ are missing.

• linnet If the argument ‘edges’ is given, then this argument now determines the ordering of the sequence of line segments. For example, the i-th row of ‘edges’ specifies the i-th line segment in as.psp(L).

• funxy, distfun The functions created by funxy and distfun have arguments (x,y). The user may now give a ppp or lpp object for the argument ‘x’, instead of giving two coordinate vectors ‘x’ and ‘y’.

• crossing.psp New argument ‘details’ gives more information about the intersections between the segments.

• subset.ppp, subset.lpp, subset.pp3, subset.ppx The argument ‘subset’ can now be any argument acceptable to the “[” method.

• density.lpp New argument ‘weights’.

• pcf.ppp New argument ‘ratio’ allows several estimates of pcf to be pooled.

• summary.ppm New argument ‘fine’ selects the algorithm for variance estimation.

• texturemap Argument ‘textures’ can be missing or NULL.

• plot.lpp New argument ‘show.network’

• linnet New argument ‘warn’

• mppm Performs more checks for consistency of the input data.

• mppm New arguments ‘gcontrol’ and ‘reltol.pql’ control the fitting algorithm.

• edge.Trans New argument ‘gW’ for efficiency.

• pool.fv The default plot of the pooled function no longer includes the variance curves.

• clickpoly The polygon is now drawn progressively as the user clicks new vertices.

• Kest Accelerated computation (for translation and rigid corrections) when window is an irregular shape.

• vcov.ppm, leverage.ppm, influence.ppm, dfbetas.ppm Performance slightly improved, for Gibbs models.

• Internal code Performance slightly improved.

• Fest, Hest Additional checks for errors in input data.

BUG FIXES

• leverage.ppm, influence.ppm, parres.ppm, addvar.ppm Calculations were completely incorrect for Gibbs models, due to a coding error. Fixed.

• update.kppm If the call to ‘update’ did not include a formula argument or a point pattern argument, then all arguments were ignored. Example: update(fit, improve.type=”quasi”) was identical to ‘fit’. Fixed.

• diagnose.ppm When applied to a model obtained from subfits(), in the default case (oldstyle=FALSE) the variance calculations were incorrect. Consequently the dotted lines representing significance bands were incorrect. An error or warning about negative variances occurred sometimes. However, calculations with oldstyle=TRUE were correct. The default has now been changed to oldstyle=TRUE for such models.

• [.lpp The local coordinate ‘seg’ was completely incorrect, when ‘i’ was a window. Fixed.

• leverage.ppm, influence.ppm, parres.ppm, addvar.ppm Crashed for Gibbs models in which the coefficient vector had length 1, such as the stationary Hardcore model. Fixed.

• subfits Crashed if the model included factor-valued spatial covariates. [Spotted by Sven Wagner] Fixed.

• subfits If the model included factor-valued spatial covariates, and if not all levels of the factor were present in each row of the data, the resulting objects were malformed and caused errors in other code. [Spotted by Sven Wagner] Fixed.

• subfits Crashed with some random-effects formulas. [Spotted by Sven Wagner] Fixed.

• improve.kppm An error message about a missing object ‘gminus1’ occurred when vcov=TRUE, fast.vcov=FALSE and type=”clik1” or “wclik1”. Fixed.

• plot.profilepl Failed with a message about a missing object ‘finite’. Fixed.

• selfcut.psp Gave an error if marks(A) was a vector rather than a data frame. [Spotted by Paul Hewson.] Fixed.

• suffstat Gave an error for point process models with Geyer interaction. Fixed.

• nncross.lpp, distfun.lpp Crashed with obscure errors if Y consisted of a single point. Fixed.

• scan.test, scanmeasure Crashed sometimes with an error message from ‘grow.mask’. Fixed.

• dppm Crashed sometimes with a message that the point pattern could not be found. [Scoping bug.] Fixed.

• mppm, profilepl Crashed, with a message about ‘SpatstatVersion’, if the ‘spatstat’ package was neither loaded nor attached. [Spotted by Markus Herrmann.] Fixed.

• qqplot.ppm Crashed sometimes when applied to a model obtained from subfits(). Fixed.

• anova.mppm Crashed sometimes with a message about mismatched coefficients. [Spotted by Sven Wagner.] Fixed.

• anova.mppm Crashed sometimes with a message about unrecognised option ‘type=”score”’. [Spotted by Sven Wagner.] Fixed.

• split.ppx Crashed if ‘f’ was not a factor. Fixed.

• idw The result was a pixel image defined in the rectangle Frame(X) instead of Window(X). Fixed.

• ppm Parameter estimates were slightly inaccurate when the model included the Geyer() interaction and the “isotropic” edge correction was used. Fixed.

• [.ppx Crashed if the number of points selected was less than 2. Fixed.

• linnet Crashed if there were no line segments at all. Fixed.

• kppm, improve.kppm Crashed if the model was stationary and improve.type != “none”. Fixed.

• as.linim.default Did not correctly handle factor-valued data. Fixed.

• texturemap Crashed if no graphical arguments were specified. Fixed.

• vcov.mppm Ignored “…” arguments. Fixed.

• Kest If ratio=TRUE and correction=c(‘border’, ‘none’) the result did not contain ratio information. Fixed.

• plot.ppp, plot.lpp Arguments ‘chars’ and ‘cols’ were ignored in some cases. Fixed.

• ppm Ignored argument ‘emend’. Fixed.

• plot.dppm Gave warnings about unrecognised argument ‘objectname’. Fixed.

• overlap.owin Sometimes returned a very small negative value, when the correct answer was 0. Fixed.