On Wed, 1 Sep 2004, michael watson (IAH-C) wrote:
> I have a data.frame, and want to perform an analysis of variance on each
> row.
Really? On one row? Makes no sense to me!
> I am currently using aov() and summary(), but what I want to do is
> perform an analysis of variance on each row and then append the F
> statistic and the p-value to the end of the row, so I can then use these
> to filter the data later on; and I want to do all this in a for loop.
>
> How can I access the F statistic and the p-value from aov()
> programmatically?
They are not there (nor are they printed). I think you may have meant
from the summary() method for aov, in which case _please_ read the
relevant help page.
Following up example(aov):
summary(npk.aov)[[1]][, 4:5]
is a data frame, with
> print.data.frame(summary(npk.aov)[[1]][, 4:5])
F value Pr(>F)
block 4.44666643 0.015938790
N 12.25873421 0.004371812
P 0.54412982 0.474904093
K 6.16568920 0.028795054
N:P 1.37829669 0.263165283
N:K 2.14597201 0.168647879
P:K 0.03119491 0.862752086
Residuals NA NA
--
Brian D. Ripley, ripley at stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595