[R] Doing mixed stock analysis with predict function using DFA

Melanie Zoelck mzoelck at hotmail.com
Sat Jan 19 14:14:58 CET 2013

Dear R Help Mailing List Members,

I have some baseline data (attached) on which I plan to run a DFA in R to find out the best predictor variables that best allow discrimination between the different stocks. I then need to apply this DFA function to the mixed data set (attached), which contains all the predictor variables of the baseline data set, but does not contain a stock variable, as their origin is unknown and I am trying to classify the mixed stock fish back to their origin stock using the DFA function generated from the baseline data. My question is whether it is possible to use the predict function with the generated DFA function and provide the mixed data set as the newdata argument within this to generate the classification, even though the mixed data set does not contain the stock variable? 

Unfortunately, the mixstock package that has recently become available is not capable of dealing with continuous variables at this time, so I am unable to use it.

I would be grateful for any guidance you may be able to provide.

Thank you,

Melanie Zölck (Zoelck)
PhD Candidate
Galway-Mayo Institute of Technology
Marine and Freshwater Research Centre
Commercial Fisheries Research Group
Department of Life Science
Dublin Road
Republic of Ireland
E-mail: mzoelck at hotmaill.com

More information about the R-help mailing list