R/model_matrix.R

In s3alfisc/fwildclusterboot: Fast Wild Cluster Bootstrap Inference for Linear Models

model_matrix <- function(object, ...) {

  #' enhanced model.matrix functionalities
  #' @param object An object of class `lm` or `felm``
  #' @param ... Other arguments
  #' @noRd
  #' @srrstats {G2.4} *Provide appropriate mechanisms to convert between
  #' different data types:* All fixed effect variables are set to
  #' type factor internally.




  UseMethod("model_matrix")
}


model_matrix.lm <- function(object, collin.rm = TRUE, ...) {
  #' Enhanced model.matrix for objects of type lm
  #' @method model_matrix lm
  #' @param object An object of class lm
  #' @param collin.rm Should collinear variables be dropped?
  #' @param ... Other arguments
  #' @noRd

  X <- model.matrix(object)
  if (collin.rm == TRUE) {
    bn <- names(na.omit(coef(object)))
    X <- X[, colnames(X) %in% bn]
  }

  X
}

model_matrix.felm <- function(object, type, collin.rm = TRUE, ...) {
  #' Enhanced model.matrix for objects of type felm
  #' @method model_matrix felm
  #' @param object An object of class felm
  #' @param collin.rm Should collinear variables be dropped?
  #' @param type 'rhs' for right-hand side variables, 'fixef' for fixed effects
  #' @param ... Other arguments
  #' @noRd
  #' @srrstats {G2.4} *Provide appropriate mechanisms to convert between
  #' different data types:* All fixed effect variables are set to
  #' type factor internally.


  dreamerr::check_arg(type, "charin(rhs, fixef)")

  if (type == "rhs") {
    mm <- model.matrix(object)
    if (collin.rm == TRUE) {
      bn <- names(na.omit(coef(object)))
      mm <- mm[, colnames(mm) %in% bn]
    }
  } else if (type == "fixef") {
    mm <- as.data.frame(object$fe)
    # make sure all fixed effect variables are factors
    #' @srrstats {G2.4}
    #' @srrstats {G2.4d} *explicit conversion to factor via `as.factor()`* Yes.
    i <- seq_along(mm)
    mm[, i] <- lapply(i, function(x) {
      factor(mm[, x])
    })
  }

  mm
}



model_matrix.fixest <-
  function(object, type, collin.rm = TRUE, ...) {
    #' Enhanced model.matrix for objects of type fixest
    #' @method model_matrix fixest
    #' @param object An object of class fixest
    #' @param type rhs lhs or fixef
    #' @param collin.rm Should collinear variables be dropped?
    #' @param ... Other arguments
    #' @noRd
    #' @srrstats {G2.4} *Provide appropriate mechanisms to convert between
    #' different data types:* All fixed effect variables are set to
    #' type factor internally.



    dreamerr::check_arg(type, "charin(rhs, fixef)")

    if (type == "rhs") {
      mm <-
        model.matrix(
          object,
          type = "rhs",
          na.rm = TRUE,
          collin.rm = collin.rm,
          as.df = TRUE
        )
    } else if (type == "fixef") {
      mm <-
        model.matrix(
          object,
          type = type,
          na.rm = TRUE,
          collin.rm = collin.rm,
          as.df = TRUE
        )
      # make sure that all fixef vars are of type factor
      #' @srrstats {G2.4}
      i <- seq_along(mm)
      mm[, i] <- lapply(i, function(x) {
        factor(mm[, x])
      })
    }

    mm
  }


#' model_matrix.ivreg <- function(object, type, collin.rm = TRUE){
#'
#'   #' Enhanced model.matrix for objects of type ivreg
#'   #' @method model_matrix fixest
#'   #' @export
#'   #' @param object An object of class fixest
#'   #' @param collin.rm Should collinear variables be dropped?
#'   #' @param ... Other arguments
#'
#'   dreamerr::check_arg(type, "charing(endo, exo)")
#'
#'   if(type == "endo"){
#'     mm <- model.matrix(
#'     object, component = "regressors", na.rm = TRUE
#'    )[, object$endogenous, drop = FALSE]
#'   } else if(type == "exo"){
#'     mm <- model.matrix(
#'     object, component = "instruments", na.rm = TRUE
#'    )[, object$exogenous, drop = FALSE]
#'   } else{
#'     stop("type needs to be 'endo' or 'exo'.")
#'   }
#'
#' }