R help - Aug 2008 - Plotting 3 way Anova

Hi
 
I'd really like to get a bar plot showing the means of my anova data. I have
looked everywhere and can only seem to find instructions for 2 way anova's.
 
I basically want to look at the mean condition of my subjects spilt by age, sex
and year (as a factor rather than a continuous variable, hence Anova and not
Ancova). and want to show it firstly as a bar graph with standard error. I have
tried using the following from Mick Crawleys R book:

barplot(tapply(x,list(y,z),mean),

beside=T,ylim=c(0,30),col=rainbow(3))

However im not sure its correct for three way, is there another code I should be
using or do I need to add an extra bit in to get all the factors?

My data is formatted as:

Condition, Age, Sex and Year. However when I plot it, all I seem to get is males
and females or adults and juveniles, when I need both for example: male adults,
male juveniles, female adults, female juveniles. what I was hoping for was
condition on the Y axis and year along the X axis and four bars for each year to
show mean ci for each sex and age group - if that makes sense.

 

Cheers

 

H


	[[alternative HTML version deleted]]

Heidi,

The interaction2wt function in the HH package does much of what you are
looking for.
You will probably need to download HH from CRAN.

interaction2wt {HH} R Documentation 

Plot all main effects and twoway interactions in a multifactor design
Description
The main diagonal displays boxplots for the main effects of each factor. The
off-diagonals
show the interaction plots for each pair of factors. The i,j panel shows the
same factors
as the j,i but with the trace- and x-factor roles interchanged. 


The second page of 
  example(interaction2wt)
shows the simple effects of each factor in the presence of the other.

In your example, "condition on the Y axis and year along the
X axis and four bars for each year to show mean ci for each sex and age
group".
I think
   interaction2wt(Y ~ year + interaction(sex, age), simple=TRUE,
data=yourdataframe)
will show what you are looking for.

Please send executable code to the entire list if there are followup
questions.


Rich