Message-ID: <66C03CD1145C95448A4D73676DD9C2ED0AF4F7@nippu1.univ.yo.oulu.fi>
Date: 2012-04-02T04:59:09Z
From: Jari Oksanen
Subject: Simper analysis with different dissimilarity index
In-Reply-To: <DUB107-W32DACD969B801C97C84945C84C0@phx.gbl>
Federico,
The simper() function uses Bray-Curtis since they were the dissimilarities that Clarke used in his original paper. Some other dissimilarities could be used as well, but *not* Jaccard. It may be that nobody has interest to change the function to accommodate for other dissimilarities, but such amendments could be added to vegan if anybody contributes a patch (and in case anybody wants to do this, the easiest source to work with may be at http://github.com/jarioksa/vegan).
By the way, Bray-Curtis and Jaccard should have identical rank-order.
Cheers, Jari Oksanen
________________________________________
From: r-sig-ecology-bounces at r-project.org [r-sig-ecology-bounces at r-project.org] on behalf of Federico Tomasetto [federicotomasetto at hotmail.it]
Sent: 02 April 2012 05:04
To: r-sig-ecology at r-project.org
Subject: [R-sig-eco] Simper analysis with different dissimilarity index
Dear list,
I used the simper function in vegan package 2.0-3 to discriminate species between native and alien plant species in New Zealand using Bray-Curtis dissimilarity index. I would like to use a different dissimilarity index (e.g. Jaccard), because I used the Jaccard index to run a metaMDS on my community data, and because, using rankindex function, the Jaccard index had the highest rank-order similarity with gradient separation (elevation).
I tried to modify the simper code but it did not work.
Any help will be more than appreciated
Regards
Federico
[[alternative HTML version deleted]]
_______________________________________________
R-sig-ecology mailing list
R-sig-ecology at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology