Please do your homework. qda is part of MASS (uncredited by you) and the
posting guide does ask you to consult that book. If you had done so you
would have found examples (start with the one on p.340)
On Sat, 3 Jul 2004, Robin Gruna wrote:
> I'm a R newbie and I have a supervised 2-class classification problem.
> To find out the best representation of my data (dim = 45). I want to
> perform LDA und QDA on the diffrent data representations to find out,
> which is best to discriminate the 2 sets. For LDA there exists a method
> plot.lda shows (in the 2 class case) a histogramm of the data, projected
> onto the linear discriminants (pleas correct me if i'm wrong...).
You have the help pages, so please do read them, carefully. There is only
one linear discriminant, and the plot method for lda can do more than
that.
In your problem there is only one quadratic discriminant, and you could do
the same thing. I'll leave you to work out why it would be a lot less
useful.
> Now to my question: Is there a similiar possibility for plotting a QDA
> object ? I know, the discriminat function is quadratic, but perhaps
> someone has in idea of to represent this problem graphically.
--
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