This function computes predictions based on fitted binary quantile models.
# S3 method for rq.bin
predict(object, newdata, na.action = na.pass,
type = "latent", grid = TRUE, ...)an rq.counts object.
an optional data frame in which to look for variables with which to predict. If omitted, the fitted values are used.
function determining what should be done with missing values in newdata. The default is to predict NA.
the type of prediction required. See details.
either a vector of breakpoints or a logical flag. If TRUE, the breakpoints are set to seq(0.05, 0.95, by = 0.05). If FALSE, the breakpoints are extracted from object$tau. At least 2 breakpoints must be provided.
not used.
a vector or a matrix of predictions.
If type = "latent" (default), the predictions are for the conditional quantiles of the latent response variable; if type = "probability", the predictions are for the conditional probabilities of the observable response. Note that type = "probability" is possible only when normalize = "last" in rq.bin call. The probabilities are predicted over a grid of values.