R/utils_replace.R

In modelsummary: Summary Tables and Plots for Statistical Models and Data: Beautiful, Customizable, and Publication-Ready

#' replace character vector entries using a named vector dictionary
#' @noRd
#'
replace_dict <- function(x, dict, interaction = FALSE) {
  for (i in seq_along(dict)) {
    x[x == names(dict)[i]] <- dict[i]
    # substitute each side of the interaction only if the coef_map doesn't
    # include interactions, which means that the user specified them the long
    # explicit way.
    if (interaction && !any(grepl(":", names(dict)))) {
      # can't use regex because of "(Intercept)" and similar user-defined arbitrary strings
      x <- gsub(
        sprintf(":%s:", names(dict)[i]),
        paste0(":", dict[i], ":"),
        x,
        fixed = TRUE
      )
      x <- gsub(
        sprintf(":%s", names(dict)[i]),
        paste0(":", dict[i]),
        x,
        fixed = TRUE
      )
      x <- gsub(
        sprintf("%s:", names(dict)[i]),
        paste0(dict[i], ":"),
        x,
        fixed = TRUE
      )
    }
  }

  return(x)
}