Continuous-Lag Spatial Markov Chains

A set of functions is provided for 1) the stratum lengths analysis along a chosen direction, 2) fast estimation of continuous lag spatial Markov chains model parameters and probability computing (also for large data sets), 3) transition probability maps and transiograms drawing, 4) simulation methods for categorical random fields.


spMC News

CHANGES IN spMC 0.3.10

BUG FIXES: o a more consistent use of PROTECT has been implemented for the function bclm in the file spMC.c (line 1552).


BUG FIXES: o the order of the macros $(LAPACK_LIBS) and $(BLAS_LIBS) was switched to be conformed with the new CRAN guidelines.

o bibentry author fields in the inst/CITATION file are now
  following the standard "Given_1 Family_1 and Given_2 Family_2".


BUG FIXES: o sim_mcs() returns more accurate probabilities, and reliable predictions.

o embed_MC() returns the transition probabilities that are now
  consistent with the tpfit() estimation methodology. An 
  additional attribute is added to deal with the counts.


BUG FIXES: o the argument "contour" in image.pemt() is now TRUE by default, and the contour plots are drawn according the rappresented probabilities.


BUG FIXES: o the example of the tpfit_me() function now provides consistent results.


BUG FIXES: o mlen() now returns more accurate results with "avg" method.

o "transCount" consider all directions between couple of points.

o the C code was fixed in order to be compatible with OpenMP 4.0.

o the deprecated which.lines() function is now replaced by 
  which_lines() in all the examples of the manual.


BUG FIXES: o an uninitialized variable was fixed in the C procedure "transCount" when dealing with OpenMP 3.1.

o unused variabels were removed from the C procedure "nsph2".

o unused values were fixed in the C procedures "nsph2" and 


NEW FEATURES: o quench() function is introduced to reproduce the optimizations algorithms, which are required for removing spatial artifacts from the simulations.

o transiogram() function computes also the log-odds standard error
  which will be used to draw confidence intervals for transition

o plot.transiogram() can draw confidence intervals around
  empirical transition probabilities.

BUG FIXES: o transiogram() has increased accuracy on non-autotransition probabilities calculation.

o pemt() has improved probabilities calculation for multiple

DEPRECATED AND DEFUNCT: o the arguments GA, optype and in functions sim_ik() and sim_ck() are defunct. Simulation optimization is performed by the quench() function.


NEW FEATURES: o pers.multi_tpfit() will produce perspective plots for theoretical multidimensional transiograms.

o perp.pemt() will produce perspective plots for 
  pseudo-empirical multidimensional transiograms.

o sim(), sim_ck(), sim_ik(), sim_mcs() and sim_path() are now
  returning an object of class "data.frame" and "spsim".

o tpfit(), tpfit_ml(), tpfit_ils(), tpfit_me(), multi_tpfit(),
  multi_tpfit_ml(), multi_tpfit_ils(), multi_tpfit_me(), mlen()
  and pemt() allows for the choice among serveral mean lengths 
  estimation method through the argument "mle".

o transiogram() can also compute empirical probabilities for
  reversible chains throught the argument "reverse".

UTILITIES: o Other robust methods are implemented for computing more reliable mean lengths. The choice can be perfrome among the averaging method ("avg"), the maximum likelihood estimator ("mlk"), trimmed average ("trm") or trimmed median ("mdn").

DEPRECATED AND DEFUNCT: o the arguments GA, optype and in functions sim_ik() and sim_ck() are deprecated. They will be considered defunct from next version release.

o tpfit(mle = FALSE) is deprecated.

o tpfit_ml(mle = FALSE) is deprecated.

o tpfit_ils(mle = FALSE) is deprecated.

o tpfit_me(mle = FALSE) is deprecated.

o multi_tpfit(mle = FALSE) is deprecated.

o multi_tpfit_ml(mle = FALSE) is deprecated.

o multi_tpfit_ils(mle = FALSE) is deprecated.

o multi_tpfit_me(mle = FALSE) is deprecated.

o mlen(mle = FALSE) is deprecated.

o pemt(mle = FALSE) is deprecated.

BUG FIXES: o The polar transformation performed by the C procedure "nsph" was not much accurate, and computationally inefficient.

o Memory bugs were fixed for C procedures "getCKPrbs", 
  "getIKPrbs", "nsph" and "transCount".

o colors correspondence between transiograms and legend was fixed
  in functions image.multi_tpfit() and image.pemt().


NEW FEATURES: o tpfit_ml(), tpfit_ils(), tpfit_me(), multi_tpfit_ml(), multi_tpfit_ils(), multi_tpfit_me() and other functions are introduced for R-methods consistency (see list of deprecated and defunct functions).

o multi_tpfit() and tpfit() may perform all the other estimation
  methods by the use of the new argument method.

o sim() may simulate and predict with all the available methods.

UTILITIES: o contour.pemt() display contours of the pseudo-empirical multidimensional and superpose the contour lines of the theoretical transition probabilities.

o image.pemt() plot the pseudo-empirical multidimensional 
  transiograms, which where previously computed.

o is.pemt() tests an object if it is of the class 'pemt'.

o pemt() returns pseudo-empirical multidimensional transiograms.

DEPRECATED AND DEFUNCT: o ck.sim() is defunct. Use sim_ck() instead.

o embed.MC() is defunct. Use embed_MC() instead.

o ik.sim() is defunct. Use sim_ik() instead.

o ilstpfit() and multi.ilstpfit() are defunct. 

  Iterated least squares are available by the use of the functions
  tpfit_ils() or multi_tpfit_ils().

o image.multi.tpfit() is defunct. Use image.multi_tpfit() instead.

o imgMultiTransiogram() is defunct.

  Pseudo-empirical multidimensional transiograms are now computed 
  first by the function pemt() and then plotted through 

o is.multi.tpfit() is defunct. Use is.multi_tpfit() instead.

o is.multi.transiogram() is defunct. Use is.multi_transiogram()

o mcs.sim() is defunct. Use sim_mcs() instead.

o metpfit and multi.metpfit() are defunct.

  Maximum entropy methods is available by the use of the functions
  tpfit_me() or multi_tpfit_me().

o multi.tpfit() is defunct.

  Multidimensional version of tpfit() is available by the use of
  the function multi_tpfit().

o path.sim() is defunct. Use sim_path() instead.

o predict.multi.tpfit() is defunct. Use predict.multi_tpfit()

o print.multi.tpfit() is defunct. Use print.multi_tpfit() instead.

o print.multi.transiogram() is defunct. Use 
  print.multi_transiogram() instead.

o which.lines() is defunct. Use which_lines().

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


0.3.11 by Luca Sartore, 5 months ago

Browse source code at

Authors: Luca Sartore [aut, cre]

Documentation:   PDF Manual  

GPL (>= 2) license

Depends on base, methods, datasets, utils, grDevices, graphics, stats

See at CRAN