R/slm.R

#' @title slm: A package for stationary linear models
#'
#' @description The \code{slm} package enables to fit linear models on datasets considering the dependence between the observations.
#'  Most of the functions are based on the functions and methods of \code{lm}, with the same arguments and the same format for the outputs.
#'
#' @section \code{slm} function, in "slm-main.R":
#'  The \code{slm} function is the main function of this package. Its architecture is the same as the \code{lm} function
#'  but it takes into account the possible correlation between the observations. To estimate the asymptotic covariance matrix of
#'  the least squares estimator, several approaches are available: "fitAR" calls the
#'  \code{cov_AR} function, "spectralproj" the \code{cov_spectralproj} function, "kernel" the \code{cov_kernel} function,
#'  "efromovich" the \code{cov_efromovich} function and "select" the \code{cov_select} function. The "hac" method uses the \code{sandwich} package,
#'  and more precisely, the method described by Andrews (1991) and Zeileis (2004).
#'
#' @section Methods for \code{slm}, in "slm-method.R":
#'  The \code{slm} function has several associated methods, which are the same as for the \code{lm} function.
#'  The available methods are: \code{summary}, \code{confint}, \code{predict}, \code{plot} and \code{vcov}.
#'
#' @section Others functions, in "auxiliary-fun.R":
#'  The package has some auxiliary functions, in particular some predefined kernels for the kernel method of \code{slm} function: the
#'  trapeze kernel, the triangle kernel and the rectangular kernel. The user can also define his own kernel and put it in the argument
#'  \code{kernel_fonc} in the \code{slm} function.
#'
#' @section Generative functions, in "generative.R":
#'  The \code{generative_process} function generates some stationary processes.
#'  The \code{generative_model} function generates some designs.
#'
#' @section Data:
#'  The package contains a dataset "shan". This dataset comes from a study about fine particle pollution in the city of Shanghai. The data are available on the
#'  following website \url{https://archive.ics.uci.edu/ml/datasets/PM2.5+Data+of+Five+Chinese+Cities#}.
#'
#' @references
#'  D. Andrews (1991). Heteroskedasticity and autocorrelation consistent covariant matrix estimation. \emph{Econometrica, 59(3), 817-858}.
#'
#'  E. Caron, J. Dedecker and B. Michel (2019). Linear regression with stationary errors: the R package slm. \emph{arXiv preprint arXiv:1906.06583}.
#'  \url{https://arxiv.org/abs/1906.06583}.
#'
#'  A. Zeileis (2004). Econometric computing with HC and HAC covariance matrix estimators.
#'
#' @docType package
#' @name slm-package
NULL

Try the slm package in your browser

Any scripts or data that you put into this service are public.

slm documentation built on Aug. 31, 2020, 5:11 p.m.