At 08:45 AM 7/12/2004, marzban wrote:
>
>Hello,
>
>For a simple problem with 1 predictor (x) and 2 classes (0 and 1), the
>linear discriminant function should be something like
>
> 2(mu_0 - mu_1)/var x + x-independent-terms
>
>where var is the common variance.
>
>Question 1: Why does lda() report only a single "Coefficients of
linear
>discriminants" when there are in fact two coefficients (the
x-dependent
>and the x-independent terms)?
>
>Question 2: And how is that single coefficient computed? It is certainly
>not equal to 2(mu_0 -mu_1)/var .
>
>Regards,
>Caren
>--
>http://www.nhn.ou.edu/~marzban
Perhaps some reading would be helpful. I suggest you look first at the
help file for lda(). Second, I suggest you read Venables and Ripley, MASS,
4th Edition, where lda() is discussed extensively. Third, I suggest you
read Ripley's Pattern Recognition and Neural Networks, where the theory is
laid out clearly. Both of these latter books are referenced in lda's help
file.
Finally, you might want to tell us what version of lda() you're using, what
version of R you're using, and what platform you're running on. For
all
we know, you're using a 2-year old version of R and lda, both long
superceded by vastly improved programs and packages.