created fortify.glm

R/fortify-glm.r

@@ -0,0 +1,56 @@
+#' Supplement the data fitted to a generalized linear model with
+#'  model fit statistics.
+#'
+#' If you have missing values in your model data, you may need to refit
+#' the model with \code{na.action = na.exclude}.
+#'
+#'
+#' @return The original data with extra columns:
+#'   \item{.hat}{Diagonal of the hat matrix}
+#'   \item{.sigma}{Estimate of residual standard deviation when
+#'     corresponding observation is dropped from model}
+#'   \item{.cooksd}{Cooks distance, \code{\link{cooks.distance}}}
+#'   \item{.fitted}{Fitted values of model}
+#'   \item{.resid}{Residuals}
+#'   \item{.stdresid}{Standardised residuals}
+#' @param model linear model
+#' @param data data set, defaults to data used to fit model
+#' @param type the type of prediction required for fitted values.
+#' See \code{link{predict.glm}}
+#' @param ... not used by this method
+#' @export
+#' @examples
+#'
+#' ## Dobson (1990) Page 93: Randomized Controlled Trial :
+#' counts <- c(18,17,15,20,10,20,25,13,12)
+#' outcome <- gl(3,1,9)
+#' treatment <- gl(3,3)
+#' d.AD <- data.frame(treatment, outcome, counts)
+#' glm.D93 <- glm(counts ~ outcome + treatment, family = poisson())
+#'
+#' head(fortify(glm.D93))
+#' ## .fitted defaults to type = "link"
+#' head(fortify(glm.D93, type = "response"))
+#'
+#' qplot(.fitted, .resid, data = mod) +
+#'   geom_hline(yintercept = 0) +
+#'   geom_smooth(se = FALSE)
+#'
+#'## Defaults make no sense to compare fitted vs actual values
+#' qplot(.fitted, counts, data = mod) +
+#'   geom_smooth(se = FALSE)
+#'
+#' qplot(.fitted, counts, data = fortify(mod, type = "response")) +
+#'   geom_smooth(se = FALSE)
+#'
+fortify.glm <- function (model, data = model$model, type = "link", ...)
+{
+  infl <- influence(model, do.coef = FALSE)
+  data$.hat <- infl$hat
+  data$.sigma <- infl$sigma
+  data$.cooksd <- cooks.distance(model, infl)
+  data$.fitted <- predict(model, type = type)
+  data$.resid <- resid(model)
+  data$.stdresid <- rstandard(model, infl)
+  data
+}