[BioC] minor fitPLM bug in get.psi.code

Paul Boutros Paul.Boutros at utoronto.ca
Sun Jul 20 05:15:09 CEST 2008


I think there's a minor bug in get.psi.code of fitPLM.  Here's the function:

> get.psi.code
function (name) 
    psi.names <- c("Huber", "fair", "Cauchy", "Geman-McClure", 
        "Welsch", "Tukey", "Andrews")
    if (!is.element(name, psi.names)) {
        stop(paste(name, "is not a valid Psi type. Please use one of:", 
            "Huber", "fair", "Cauchy", "Geman-Mclure", "Welsch", 
            "Tukey", "Andrews"))
    code <- c(0:6)[name == psi.names]

In psi.names function is checking for the string "Geman-McClure" but in the
stop statement it reports the name as "Geman-Mclure" (missing a C).


More information about the Bioconductor mailing list