fnH2, Fractional Brownian Motion (FBM) kernel fnH3 (with a
default Hurst coefficient of 0.5), and the Pearson kernel fnH1.
fnH1(x, y = NULL)
fnH2(x, y = NULL)
fnH3(x, y = NULL, gamma = 0.5)x and y must have
similar dimensions."Canonical", "FBM,gamma",
or "Pearson" whose [i, j] entries are $h($y[i],
x[j]$)$, with $h$ being the kernel function. The matrix has
dimensions m by n according to the lengths of y and
x which has lengtsh m and n respectively. When a
single vector argument x is supplied, then y is taken to be
equal to x, and a symmetric n by n matrix is returned.If x is a matrix or data frame with p columns, then the
kernel matrix returned is fnH(x[, 1]) + ... + fnH(x[, p]).
factor-type objects are treated with the Pearson kernel
automatically. The other two kernel types are for continuous variables, with
the Canonical kernel used for "straight-line" effects and the FBM for
smoothing effects. The smoothness is controlled somewhat by the Hurst
coefficient.