[R] visualize TukeyHSD results

Mendiburu, Felipe (CIP) F.MENDIBURU at CGIAR.ORG
Tue Nov 2 20:44:59 CET 2010

Dear Timothy,

Use library(agricolae)
> library(agricolae)
> a = aov(Weight~Feed)
> HSD.test(a,"Feed")

HSD.test(a,"Feed", group=TRUE)
HSD.test(a,"Feed", group=FALSE)


Felipe de Mendiburu.
International Potato Center. www.cipotato.org
University: Agraria La Molina - Peru. www.lamolina.edu.pe

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Timothy Spier
Sent: Thursday, October 21, 2010 9:50 PM
To: r-help at r-project.org
Subject: [R] visualize TukeyHSD results

I am a new R user but a long time SAS user. I searched for a response to
this question but no luck, so forgive me if this topic has been covered
before. I am running a TukeyHSD post hoc test after running an ANOVA. I
get the results of all pairwise comparisons, no problem. However, the
output table is a little "busy", and I'd like to make the output easier
to read. Specifically, I would like all groups which are not
significantly different to be given the same letter. 

For example, here is a simple ANOVA with Tukey post hoc. It compares
weight gain in pigs among 4 feeds labeled "A", "B", "C", and "D":

> a = aov(Weight~Feed)
> TukeyHSD(a)
  Tukey multiple comparisons of means
    95% family-wise confidence level

Fit: aov(formula = Weight ~ Feed)

      diff        lwr       upr     p adj
B-A   6.68   1.096263 12.263737 0.0168421
C-A   8.73   2.807553 14.652447 0.0034914
D-A  -1.38  -6.963737  4.203737 0.8906642
C-B   2.05  -3.872447  7.972447 0.7530266
D-B  -8.06 -13.643737 -2.476263 0.0041505
D-C -10.11 -16.032447 -4.187553 0.0009497

What I really want would look something like this:

Feed         Mean     TukeyResult
C            73.4           a
B            71.3           a
A            64.6           b
D            63.2           b

Any ideas?
	[[alternative HTML version deleted]]

R-help at r-project.org mailing list
PLEASE do read the posting guide
and provide commented, minimal, self-contained, reproducible code.

More information about the R-help mailing list