R/splitpar.R

In mlesnoff/rnirs: Dimension reduction, Regression and Discrimination for Chemometrics

splitpar <- function(x, a1 = .1, a2 = 10, n = 20) {
    
    lim <- x * c(a1, a2)
    
    z1 <- seq(lim[1], lim[2], length.out = round(n / 2))
    z2 <- exp(seq(log(lim[1]), log(lim[2]), length.out = round(n / 2)))
    z2 <- z2[-c(1, length(z2))]
    z3 <- seq(.8, 1.2, by = .1) * x
    z3 <- z3[z3 >= lim[1] & z3 <= lim[2]]
    
    sort(unique(round(c(z1, z2, z3), digits = 10)))
    
    }