1\name{rq.object} 2\alias{rq.object} 3\alias{formula.rq} 4\alias{logLik.rq} 5\alias{logLik.rqs} 6\alias{AIC.rq} 7\alias{AIC.rqs} 8\alias{extractAIC.rq} 9\title{ 10Linear Quantile Regression Object 11} 12\description{ 13 These are objects of class \code{"rq"}. 14 They represent the fit of a linear conditional quantile function model. 15} 16\section{Generation}{ 17 This class of objects is returned from the \code{rq} function 18 to represent a fitted linear quantile regression model. 19} 20\section{Methods}{ 21 The \code{"rq"} class of objects has methods for the following generic 22functions: 23\code{coef}, \code{effects} 24, \code{formula} 25, \code{labels} 26, \code{model.frame} 27, \code{model.matrix} 28, \code{plot} 29, \code{logLik} 30, \code{AIC} 31, \code{extractAIC} 32, \code{predict} 33, \code{print} 34, \code{print.summary} 35, \code{residuals} 36, \code{summary} 37} 38\section{Structure}{ 39 The following components must be included in a legitimate \code{rq} object. 40 \describe{ 41 \item{\code{coefficients}}{ 42 the coefficients of the quantile regression fit. 43 The names of the coefficients are the names of the 44 single-degree-of-freedom effects (the columns of the 45 model matrix). 46 If the model was fitted by method \code{"br"} with \code{ci=TRUE}, then 47 the coefficient component consists of a matrix whose 48 first column consists of the vector of estimated coefficients 49 and the second and third columns are the lower and upper 50 limits of a confidence interval for the respective coefficients. 51 } 52 \item{\code{residuals}}{ 53 the residuals from the fit. 54 } 55 \item{\code{dual}}{ 56 the vector dual variables from the fit. 57 } 58 \item{\code{rho}}{ 59 The value(s) of objective function at the solution. 60 } 61 \item{\code{contrasts}}{ 62 a list containing sufficient information to construct the contrasts 63 used to fit any factors occurring in the model. 64 The list contains entries that are either matrices or character vectors. 65 When a factor is coded by contrasts, the corresponding contrast matrix 66 is stored in this list. 67 Factors that appear only as dummy variables and variables in the model 68 that are matrices correspond to character vectors in the list. 69 The character vector has the level names for a factor or the column 70 labels for a matrix. 71 } 72 \item{\code{model}}{ 73 optionally the model frame, if \code{model=TRUE}. 74 } 75 \item{\code{x}}{ 76 optionally the model matrix, if \code{x=TRUE}. 77 } 78 \item{\code{y}}{ 79 optionally the response, if \code{y=TRUE}. 80 } 81 } 82} 83\details{ 84 The coefficients, residuals, and effects may be extracted 85 by the generic functions of the same name, rather than 86 by the \code{$} operator. For pure \code{rq} objects this is less critical 87 than for some of the inheritor classes. In particular, for fitted rq objects 88 using "lasso" and "scad" penalties, \code{logLik} and \code{AIC} functions 89 compute degrees of freedom of the fitted model as the number of estimated 90 parameters whose absolute value exceeds a threshold \code{edfThresh}. By 91 default this threshold is 0.0001, but this can be passed via the \code{AIC} 92 function if this value is deemed unsatisfactory. The function \code{AIC} 93 is a generic function in R, with parameter \code{k} that controls the form 94 of the penalty: the default value of \code{k} is 2 which yields the classical 95 Akaike form of the penalty, while \code{k <= 0} yields the Schwarz (BIC) 96 form of the penalty. 97 Note that the extractor function \code{coef} returns a vector with missing values 98 omitted. 99} 100\seealso{ 101 \code{\link{rq}}, \code{\link{coefficients}}. 102} 103\keyword{regression} 104