[R] Presence Absence of variables in a multivariate data set

I'm sure this is a simple problem, however I need a bit of help to sort it out. I have a multivariate data set of elements. Some of the observations have some amount of certain elements but are lacking in others. Therefore, I want to quickly group/cluster the observations based on a presence/ab

[R] predict.lda posteriors equal to 1

Hi all, I was wondering if anyone was able to explain to me why when I run predict.lda() on some new data I get lots of posterior probabilities = 1. This seems markedly unrealistic. What do the posteriors mean in the context of the predict.lda() function? Thanks Gareth -- View this message i

[R] predict.lda posteriors equal to 1

Hi all, I was wondering if anyone was able to explain to me why when I run predict.lda() on some new data I get lots of posterior probabilities = 1. This seems markedly unrealistic. What do the posteriors mean in the context of the predict.lda() function? Thanks Gareth -- View this message i

[R] mtext problem - grainy and massive file size

Hi Team, I'm on a mac, 0SX 10.5.6, R 2.7.2. I consistently make pairs plots and use mtext() to put labels on top of the plot in the outer margin. All of a sudden, this text has become really grainy in the final product and is making huge files. I have no idea why this has just started happenin

Re: [R] Variable Selection for data reduction and discriminant anlaysis

Thanks Mark, I failed to mention that i'm working within a compositional framework. I didn't want to confuse things. My data is transformed to the clr or alr under Aitchison geometry, so I am essentially working in Euclidean space. Has anyone had experience doing stepwise LDA?? I can't for t