denpro-package {denpro} | R Documentation |
The package provides tools to (1) visualize multivariate density functions and density estimates with level set trees, (2) visualize level sets with shape trees, (3) visualize multivariate data with tail trees, (4) visualize scales of multivariate density estimates with mode graphs and branching maps, and (5) visualize anisotropic spread with 2D volume functions and 2D probability conetent functions. With level set trees one visualizes mode structure, with shape trees one visualizes shapes of level sets of unimodal densities, and with tail trees one visualizes connected data sets. The kernel estimator is implemented but the package may be applied for visualizing other density estimates, which have to be setwise constant.
Package: | denpro |
Version: | 0.8.2 |
Date: | 2006-05-16 |
Depends: | R |
License: | GPL version 2 or newer |
URL: | http://www.rni.helsinki.fi/~jsk/denpro |
Packaged: | Thu May 18 15:34:27 2006; jsk |
Built: | R 2.2.1; i686-pc-linux-gnu; 2006-05-18 15:34:43; unix |
The main function:
*leafsfirst
Functions to plot a level set tree, shape tree, or tail tree:
*plotvolu *plotbary *plottree *treedisc *prunemodes
Visualization of scales of estimates:
* lstseq.kern * modegraph * plotmodet * branchmap * plotbranchmap * scaletable * exmap * plotexmap * hgrid :
Visualization of anisotropic spread (2D volume function and 2D probability content function):
* stseq * shape2d * plotvolu2d * plotdelineator
Other utilities:
* pcf.kern * tree.segme * lst2xy * graph.matrix * plotbary.slide * Calculation of level set trees with "DynaDecompose" algorithm and with naive algorithm o profkern o profhist o profgene o profeval * locofmax * modecent * excmas
Other visualization tools:
* draw.pcf * slicing * draw.levset * paracoor * pp.plot * qq.plot * plot.kernscale * dist.func * paraclus
Miscallenous:
* pcf.func * sim.data * dend2parent * explo.compa * nn.radit * nn.likeset * rotation * liketree * leafsfirst.visu * plottwin
Index (alphaphetical):
branchmap Calculates a branching map from a sequence of level set trees draw.levset Plots a level set of a 2D function draw.pcf Prepares the plotting of a 2D or 1D piecewise constant function excmas Calculates the excess masses associated with the nodes of a level set tree exmap Calculates a scale of excess mass profiles hgrid Returns a grid of smoothing parameter values leafsfirst Calculates a level set tree, shape tree, or tail tree locofmax Calculates the location of the maximum of a function lstseq.kern Calculates a scale of kernel estimates modecent Returns locations of modes of a density estimate modegraph Calculates a mode graph from a scale of estimates paracoor Makes a parallel coordinate plot pcf.func Calculates a piecewise constant function for some illustrative purposes pcf.kern Calculates a multivariate kernel estimate plotbary Makes a barycenter plot of a level set tree, a location plot of a shape tree, or a tail tree plot of a tail tree plotbranchmap Plots a branching map plotexmap Plots a scale of excess mass profiles plotmodet Plots a mode graph plottree Makes a tree plot of a level set tree, of a shape tree, or of a tail tree plotvolu Makes a volume plot of a level set tree, a shape plot of a shape tree, or a tail frequency plot of a tail tree plotvolu2d Makes a perspective plot of a 2D volume function or a 2D probability content function profgene Calculates the level set tree of a rectangularwise constant function profhist Calculates the level set tree of a histogram profkern Calculates a level set tree of a kernel estimate prunemodes Prunes modes away from a level set tree or a shape tree scaletable Implements the scale and shape table shape2d Returns a 2D volume function or 2D probability content function sim.data Generates data for illustrative purposes slicing Returns a one- or two-dimensional slice of a multivariate function stseq Calculates a sequence of radius functions from a sequence of level sets treedisc Prunes a level set tree or a tail tree tree.segme Returns the segmentation of the nodes of a visualization tree
Jussi Klemela <klemela@oulu.fi>
Maintainer: Jussi Klemela <klemela@oulu.fi>
http://www.rni.helsinki.fi/~jsk/denpro
# level set tree dendat<-sim.data(n=200,type="mulmod") # data pcf<-pcf.kern(dendat,h=1,N=c(32,32)) # kernel estimate lst<-leafsfirst(pcf) # level set tree td<-treedisc(lst,pcf,ngrid=60) # pruned level set tree plotvolu(td) # volume plot plotbary(td) # barycenter plot # shape tree dendat<-sim.data(n=200,type="cross") # data pcf<-pcf.kern(dendat,h=1,N=c(32,32)) # kernel estimate st<-leafsfirst(pcf,propo=0.01) # shape tree, 1 per cent level set tdst<-treedisc(st,pcf,ngrid=60) # pruned shape tree plotvolu(tdst) # radius plot plotbary(tdst) # location plot # tail tree dendat<-sim.data(n=200,type="cross") # data tt<-leafsfirst(dendat=dendat,rho=0.65) # tail tree plotbary(tt) # tail tree plot plotvolu(tt) # tail frequency plot