R/B.r

In logbin: Relative Risk Regression Using the Log-Binomial Model

B <- function(..., knots = NULL, knot.range = 0:5) {
  vars <- as.list(substitute(list(...)))[-1]
  if(length(vars) == 0) 
    stop("B() is not supported")
  if(length(vars) > 1)
    stop("B() with more than one variable is not supported")
  if(is.null(knots) && is.null(knot.range))
    stop("either 'knots' or 'knot.range' must be specified with B()")
    
  knots <- as.vector(knots)
  if(!is.null(knots)) {
    if(!is.numeric(knots)) stop("knots must be numeric")
    if(length(knots) != length(unique(knots))) stop("knots must be unique")
    knots <- sort(knots)
    knot.range <- length(knots)
  }

  term <- deparse(vars[[1]], backtick = TRUE, width.cutoff = 500)
  if (term[1] == ".")
    stop("B(.) is not supported")
  termname <- attr(terms(reformulate(term[1])),"term.labels")
  termlabel <- paste0("B(", termname, ")")
  ret <- list(term = term, termlabel = termlabel, knots = knots, 
              knot.range = knot.range)
  class(ret) <- "B.smooth"
  ret
}