R/summary.slmfit.R

In highamm/sptotal: Predicting Totals and Weighted Sums from Spatial Data

#' Summarizes a fitted spatial linear model.
#'
#' In conjunction with \code{print.summary.slmfit()}, the output looks similar
#' to output from \code{R}'s standard \code{lm()} function.
#'
#' @param object is an object generated from \code{\link{slmfit}()} of
#' class \code{slmfit}.
#' @param ... further arguments passed to or from other methods.
#' @return a list with \itemize{
#'   \item model formula
#'   \item a table of fixed effects estimates and associated standard errors
#'   \item estimated spatial covariance parameter estimates
#'   \item residuals
#'   \item generalized r-squared.
#'        }
#' @examples
#' data(exampledataset) ## load a toy data set
#' slmobj <- slmfit(formula = counts ~ pred1 + pred2, data = exampledataset,
#' xcoordcol = 'xcoords', ycoordcol = 'ycoords', areacol = 'areavar')
#' summary(slmobj)
#' @export

summary.slmfit <- function(object, ...) {

  catcall <- object$FPBKpredobj$formula

  predictornames <- object$PredictorNames
  NAvec <- rep(NA, times = length(predictornames))

  regcoefs <- object$CoefficientEsts
  regvar <- object$BetaCov
  p <- length(regcoefs)
  n <- object$SampSize

  sereg <- sqrt(diag(as.matrix(regvar)))

  tvec <- NAvec
  tvec <- regcoefs / sereg
  pvec <- NAvec
  pvec <- round(100000 * (1 - stats::pt(abs(regcoefs / sereg),
    df = n - p)) * 2) / 100000

  fixed.eff.est <- data.frame(##FactorLevel = predictornames,
    Estimate = regcoefs,
    std.err = sereg, t.value = tvec, prob.t = pvec)
  fixed.effects.estimates <- fixed.eff.est

  covmodels <- object$SpatialParmEsts
  covmodelout <- data.frame(covmodels)
  colnames(covmodelout) <- paste(object$CovarianceMod, "Model")

  resid_vec <- object$resids
  ##residualsum <- c(min(residuals), quantile(residuals, c(0.25, 0.5,
  ##  0.75)), max(residuals))
  generalizedr2 <- GR2(object)

  outpt <- list(catcall = catcall,
    fixed.effects.estimates = fixed.effects.estimates,
    covariance.parameters = covmodelout,
    resid_vec,
    generalizedr2)
  names(outpt) <- c("catCall", "FixedEffects", "CovarianceParms",
    "Residuals", "GeneralizedR2")
  class(outpt) <- "summary.slmfit"
  return(outpt)

}