Linear Discrimination

Finds the coefficients \(a_k\) and constants \(c_k\) for Fisher's linear discrimination function \(d_k\) in (11.31) and (11.32).

lda(x, y)

Arguments

x

The \(N \times P\) data matrix.

y

The \(N\)-vector of group identities, assumed to be given by the numbers 1,...,\(K\) for \(K\) groups.

Value

A list with the following components:

a

A \(P x K\) matrix, where column \(K\) contains the coefficents \(a_k\) for (11.31). The final column is all zero.

c

The \(K\)-vector of constants \(c_k\) for (11.31). The final value is zero.

See also

sweep

Examples

# Iris example
x.iris <- as.matrix(iris[, 1:4])
# Gets group vector (1, ... , 1, 2, ... , 2, 3, ..., 3)
y.iris <- rep(1:3, c(50, 50, 50))
ld.iris <- lda(x.iris, y.iris)