I have a problem in implementing an analysis of variance model in S+. The
data are set in a data frame, with a column indicating the treatment (2
levels), a column indicating the gene to which treatments were applied (2
rows by gene), and 8 columns giving repeated measurement over time, not
equally spaced.
As an example :
time0 time6 time12 time24 time30 time36 time50
time54 genes treatment
9400.333 10616.333 15263.000 13249.333 10789.000 10761.667 11073.000
11388.000 1 1
11278.333 6976.667 5455.667 11885.000 10211.667 11717.333 10777.000
14611.667 1 2
7022.667 8465.000 4073.000 6755.667 6790.333 8860.333 9485.667
7007.333 2 1
6368.000 8368.000 9889.000 8654.000 7171.667 8297.667 6425.667
8096.667 2 2
5621.000 3423.000 4127.000 4669.667 4636.333 6453.667 6802.667
6290.333 3 1
7706.000 4364.000 3751.333 6084.333 5295.000 7130.667 5550.667
6590.000 3 2
>
............................................................................
.................
The problem is that I don't know how to write the proper S+ code for that
kind of analysis (including treatment effect, gene effect, and time effect).
Browsing books didn't give me a clear answer.
Thanks if someone can help.
Andre MERY
Biometry Department
Aventis Pharma France
andre.mery@aventis.com
|