[R] Create grouping from TukeyHSD (as a duncan test does)?

[Van Dongen Stefan]

Hi Bernhard,

When you are interested in so many groups and to compare them, why not use a cluster technique or a mixture approach to get some grouping

Stefan

-----Oorspronkelijk bericht-----
Van: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] Namens bernhard.mueller at art.admin.ch
Verzonden: dinsdag 25 september 2007 14:31
Aan: r-help at r-project.org
Onderwerp: [R] Create grouping from TukeyHSD (as a duncan test does)?


Hello everybody

1. If there is/ever will be a function to perform "duncan multiple range test" please inform me at once.

2. Is there a way to create a grouping as duncan does from TukeyHSD output? My experimental design contained 62 genotypes, so the pairwise comparison is not that usefull and clear to look at. How would I do that? Is there an other test that does give grouping as output?

Thnaks for your help.

Bernhard Müller, Master Student
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Eidgenössisches Volkswirtschaftsdepartement EVD
Forschungsanstalt Agroscope Reckenholz-Tänikon ART
Ökologische Landbausysteme

Reckenholzstrasse 191,  CH-8046 Zürich
Tel.    ++41 (0)44 / 377 74 47
FAX     ++41 (0)44 / 377 72 01
mailto:bernhard.mueller at art.admin.ch
http://www.reckenholz.ch


	[[alternative HTML version deleted]]