[R-sig-eco] predict NMS scores for new samples

[Jonathan Coop]

hi all,

I'm wondering if anyone has developed code to predict NMS scores for new sample data, as the NMDS Scores Prediction algorithm functioned in PC-Ord.  I am looking at the effects of multiple fires (reburns) on vegetation composition.  I used an NMS ordination of pre-reburn vegetation samples to characterize vegetation, examine relationships between vegetation and subsequent fire severity, and stratify a re-sampling effort by key vegetation "types".  I'd love to go back now with the post-reburn vegetation samples and see where they fall out in the original ordination space.

Any and all thoughts welcome!  

Thanks,

Jonathan Coop



 		 	   		  
	[[alternative HTML version deleted]]