R/lsinf.R

In Auburngrads/SMRD: Statistical Methods For Reliability Data

#' Title
#'
#' @param z 
#' @param censor.type 
#' @param distribution 
#' @param force 
#' @param debug1 
#'
#' @return list of values
#' @export
#'
#' @examples
#' \dontrun{
#' 
#' lsinf(seq(-1,1, by = 0.1),"right","sev")
#' 
#' lsinf(seq(-2,2, by = 0.2),"right","normal")
#' 
#' }
lsinf <-
function (z, 
          censor.type, 
          distribution, 
          force = F,
          debug1 = F)
{
    if (!is.numeric(z)) stop("z must be numeric")
    if (!is.character(censor.type)) stop("censor.type must be character")
    if (!is.character(distribution)) stop("distribution must be character string")
    
    switch(generic.distribution(distribution), 
           weibull = idist <- 2, 
           sev = idist <- 1,
           frechet = , lev = idist <- 2, 
           lognormal = , normal = idist <- 3,
           loglogistic = , logistic = idist <- 4, 
           stop("Distribution must be sev, lev, normal, or logistic"))
    switch(censor.type, 
           uncensored = icode <- 1, 
           right = icode <- 2,
           left = icode <- 3, 
           stop("censor.type must be uncensored, left, or right"))
    nrows <- length(z)
    if (debug1) browser()
    
    zout <- SLSINF(as.integer(idist),
                   as.integer(icode),
                   as.double(z),
                   as.double(z),
                   f11 = double(nrows),
                   f12 = double(nrows),
                   f22 = double(nrows),
                   as.integer(nrows),
                   ifault = integer(1),
                   irow = integer(1))
    
    if(length(z)==1) {
        
       zmat <- matrix(c(zout$f11,zout$f12,zout$f12,zout$f22), nrow = 2, byrow = TRUE)
       colnames(zmat) <- c("f_i1", "f_i2")
       rownames(zmat) <- c("f_1j", "f_2j")
       
    } else {
        
      zmat <- NULL
      
    }
    
    return(list(f11 = zout$f11, f12 = zout$f12, f22 = zout$f22, matrix = zmat))
}