R/sumz.R

In metap: Meta-Analysis of Significance Values

sumz <-
function(p, weights = NULL, data = NULL, subset = NULL,
   na.action = na.fail, log.p = FALSE, log.input = FALSE)  {
   if(is.null(data)) data <- sys.frame(sys.parent())
   mf <- match.call()
   mf$data <- NULL
   mf$subset <- NULL
   mf$na.action <- NULL
   mf[[1]] <- as.name("data.frame")
   mf <- eval(mf, data)
   if(!is.null(subset)) mf <- mf[subset,]
   mf <- na.action(mf)
   p <- as.numeric(mf$p)
   weights <- mf$weights
   noweights <- is.null(weights)
   if(noweights) weights <- rep(1, length(p))
   if(length(p) != length(weights)) warning("Length of p and weights differ")
   if(log.input) {
      keep <- p < 0
   } else {
      keep <- (p > 0) & (p < 1)
   }
   invalid <- sum(1L * keep) < 2
   if(invalid) {
      warning("Must have at least two valid p values")
      res <- list(z = NA_real_, p = NA_real_,
         validp = p[keep], weights = weights)
   } else {
      if(sum(1L * keep) != length(p)) {
         warning("Some studies omitted")
         omitw <- weights[!keep]
         if((sum(1L * omitw) > 0) & !noweights)
            warning("Weights omitted too")
      }
      zp <- (qnorm(p[keep], lower.tail = FALSE, log.p = log.input) %*%
         weights[keep]) / sqrt(sum(weights[keep]^2))
      res <- list(z = zp, p = pnorm(zp, lower.tail = FALSE,
            log.p = log.p),
         validp = p[keep], weights = weights)
   }
   class(res) <- c("sumz", "metap")
   res
}
print.sumz <- function(x, ...) {
   cat("sumz = ", x$z, "p = ", x$p, "\n")
   invisible(x)
}