getRhos: Compute the penalty parameters corresponding to degrees of...

In hypergsplines: Bayesian model selection with penalised splines and hyper-g prior

Description

Compute the penalty parameters corresponding to degrees of freedom

Usage

  getRhos(lambdas, degrees)

Arguments

lambdas	the eigenvalues of Z^T * W * Z, where Z is the spline basis matrix
degrees	the vector of degrees of freedom

Value

the vector of rho parameters corresponding to the degrees

Note

All degrees must be positive, and smaller than the dimension of the spline basis (which is the number of columns of Z).

Author(s)

Daniel Sabanes Bove [e-mail]

Examples

x <- seq(from=-5,
         to=5,
         length=100)
Z <- makeBasis(x=x,
               nKnots=4L)
lambdas <- svd(Z, nu=0L, nv=0L)$d^2
getRhos(lambdas=lambdas,
        degrees=c(1, 2, 3, 3.9))

hypergsplines documentation built on May 2, 2019, 6:14 p.m.