Hello everyone, I'm working with R 2.4.1 on a PC running with XP. Trying to run isoMDS as follows: Gquad.mat <- Gquads[4:10] # extracts only the metric data variables Gquad.dist <- dist(Gquad.mat) Gquad.mds <- isoMDS(Gquad.dist) plot(Gquad.mds$points, type = "n") text(Gquad.mds$points, labels = as.character(Gquads$Quadname), cex = .5) The data.frame includes 711 records, and the first 3 columns contain categorical variables which I exclude from the analysis. When I invoke the script above an error message is returned. Error in isoMDS(Gquad.dist) : zero or negative distance between objects 179 and 180. How can I handle this, is it valid to add 0.5 to every element in the distance matrix or is someother alternative more appropriate? Thanks in advance Steve Steve Friedman, PhD Everglades Division Senior Environmental Scientist, Landscape Ecology South Florida Water Management District 3301 Gun Club Road West Palm Beach, Florida 33406 email: sfriedma@sfwmd.gov Office: 561 - 682 - 6312 Fax: 561 - 682 - 5980 If you are not doing what you truly enjoy its your obligation to yourself to change. [[alternative HTML version deleted]]
>Error in isoMDS(Gquad.dist) : zero or negative distance between objects >179 and 180. >How can I handle this, is it valid to add 0.5 to every element in the >distance matrix or is someother alternative more appropriate?This means that your objects 179 and 180 are identical so just remove one of the two. They both have the same distance to all other objects so your results shouldn't change. Whatever conclusion you draw from the location of one on your plot, you can say the same for the other. Silvia. Thanks in advance Steve Steve Friedman, PhD Everglades Division Senior Environmental Scientist, Landscape Ecology South Florida Water Management District 3301 Gun Club Road West Palm Beach, Florida 33406 email: sfriedma at sfwmd.gov Office: 561 - 682 - 6312 Fax: 561 - 682 - 5980 If you are not doing what you truly enjoy its your obligation to yourself to change. [[alternative HTML version deleted]] ______________________________________________ R-help at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. -- View this message in context: http://www.nabble.com/Non-metric-multidimensional-scaling-tf4489161.html#a12803598 Sent from the R help mailing list archive at Nabble.com.