Message-ID: <BAY172-W18FD0921B4538EDE98375DC5950@phx.gbl>
Date: 2014-10-22T01:05:37Z
From: Jonathan Coop
Subject: predict NMS scores for new samples
In-Reply-To: <B2167C04-3601-45F3-9C77-EE29926B4AD6@oulu.fi>
Thanks to all for the useful suggestions -- I just wanted to report back on my experience using the code that Dave Roberts refers to below, on his web page. It worked beautifully.
I imagine that the utility of this approach would depend on the compositional overlap of the new samples with the old. In the case of the research problem I am working on, there is quite a lot of overlap, and re-burning largely reinforces compositional shifts related to previous fires.
Jonathan
> From: jari.oksanen at oulu.fi
> To: droberts at montana.edu
> Date: Sun, 12 Oct 2014 20:38:53 +0000
> CC: r-sig-ecology at r-project.org
> Subject: Re: [R-sig-eco] predict NMS scores for new samples
>
>
> On 12/10/2014, at 23:13 PM, Dave Roberts wrote:
>
> > Hi Jonathan,
> >
> > If you're using the metaMDS function in vegan with the monomds engine then it's possible. I have posted a function (monomds) at the bottom of
> >
> > http://ecology.msu.montana.edu/labdsv/R/labs/lab9/lab9.html
> >
> > that shows how to generate a labdsv:::nmds object from vegan's monomds function (courtesy of Peter Minchin). You will have to have package vegan loaded to get the monomds FORTRAN code loaded.
> > Then you can use the function addpoints.nmds (also at the bottom of that page) to add points to an existing nmds.
> >
> > I'm not convinced that it's a good idea, but I worked with someone who needed it to fulfil contract obligations so I wrote it.
> >
> It looks like a very good idea to me.
>
> Cheers, Jari Oksanen
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
[[alternative HTML version deleted]]