R-users: My problem: the package fdim (which measures fractal dimensions of datasets) apparently prefers 3-D (XYZ) datasets. I'm trying to measure the information dimension (a type of fractal dimension) of a 2-D (XY) dataset. Here's some examples of the preferred 3-D data (XYZ): library(fdim) XYZ <- (makehplane(NumN=300, DimM=3, Longit=300)) fdim(XYZ,q=0.999999,PlotF=TRUE) Here's an example of the type of data I have: library(fdim) X<- sample(1:1000,size=300,replace=TRUE) Y<- sample(1:500,size=300,replace=TRUE) XY<- matrix(c(X,Y),nrow=300) fdim(XY,q=0.999999,PlotF=TRUE) My question is this: since fdim prefers XYZ data, is it invalid to use fdim to measure the fractal dimension of XY data? What I have is XY position data of individual trees in a section of forest. I'd like to measure the information dimension of the distribution of trees. The reason I'm asking my question is that fdim returns consistently higher dimension numbers than do other fractal-measuring programs. However, those other programs measure the fractal dimension of bitmaps, not datasets, and have a problem with resolution--many of the points on the bitmap overlap, making it impossible to measure the dimension of those points. Therefore, I'd like to measure the information dimension using a program that uses datasets, not bitmaps. Thanks for your input. Jim Milks Graduate Student Environmental Sciences Ph.D. Program Wright State University Dayton, OH 45435 [[alternative HTML version deleted]]