TABLE OF CONTENTS


splineUtils/evalEinte [ Functions ]

NAME

    evalEinte --- compute the expected distance from 1 of a B-spline

FUNCTION

Since the frailty density must have mean 1, it is important to be able to compute the difference between 1 and the frailty density mean. This function calculates the expected values of a set of normalized B-spline basis functions and subtracts them from 1. The difference between 1 and the frailty mean is then

       curve$spline.basisexp%*%exp(curve$spline.par)

where the former is the output of this function and the latter is the set of spline weights.

SYNOPSIS

1511 evalEinte <- function(knots, ord)

INPUTS

    knots      a set of spline knots as output by makeknots
    ord        integer spline order

OUTPUTS

    a vector containing one minus the expectation of each of the normalized basis functions

SOURCE

1514 {
1515     K <- sum(knots > attr(knots, "b")[1] & knots < attr(knots, "b")[2])
1516     Einte <- rep(0, K + ord)
1517     for(j in 1:(K + ord)){
1518         # Einte[j] contains the 1st moment of the j-th spline of order ord defined
1519         # on a given set of knots
1520         Einte[j] <- nBsmom(1, ord, j, knots)
1521     }
1522     return(1 - Einte)
1523 }