R/line_plot.R

In swehip/slrplotfun: Plot functions for the Swedish Arthroplasty Register

#' Line plot function
#'
#' Standard line plot using ggplot2. Y-variable not necessary.
#'
#' @param df               Data frame.
#' @param x_var            Variable for x-axis, use string name.
#'                           Recommended that x_var is
#'                           in character in df (not necessary).
#' @param color_var        Variable for the different colors in lines, use string
#'                           name. Use `NULL` if only one color for lines.
#' @param y_var            Variable for y axis, if `NULL`, count is used.
#' @param group_by_x_var   Boolean indicating if percentages should be for `x_var`
#'                           or `color_var`.
#' @param y_percent        If `TRUE`, y-axis is in percent form. Otherwise in count
#'                           form.
#' @param percent_accuracy Set accuracy for [scales::percent_format()].
#' @param y_lim            Limit on y-axis.
#' @param y_breaks         Length between each break on y-axis.
#' @param y_breaks_end     Break end, default for 100,000. Works for all count
#'                           values below that.
#' @param expand           If `TRUE`, the margins around the data are kept.
#' @param line_size        Size of the lines.
#' @param title            Plot title, `NULL` if no title.
#' @param subtitle         Small text under title, `NULL` if no subtitle.
#' @param y_lab            Y-axis label, use `NULL` for no label.
#' @param x_lab            X-axis label, use `NULL` for no label.
#' @param fill_colors      Colors of the different categories in color_var.
#' @param legend_labels    Label for each legend key.
#' @param label_breaks     Order of the legend keys.
#' @param legend_row       How many rows for the legends.
#' @param legend_col       How many columns for the legends.
#' @param ...              Arguments passed to [theme_slr()]
#'
#' @return                 Ggplot object containing line-plot.
#' @example                man/examples/line_plot.R
#' @export
line_plot <-
  function(
    df,
    x_var,
    color_var         = NULL,
    y_var             = NULL,
    group_by_x_var    = TRUE,
    y_percent         = TRUE,
    percent_accuracy  = 1,
    y_lim             = NULL,
    y_breaks          = 2000,
    y_breaks_end      = 100000,
    line_size         = 1,
    title             = NULL,
    subtitle          = NULL,
    y_lab             = NULL,
    x_lab             = NULL,
    fill_colors       = NULL,
    legend_labels     = ggplot2::waiver(),
    label_breaks      = ggplot2::waiver(),
    legend_row        = NULL,
    legend_col        = NULL,
    expand            = TRUE,
    ...
  ) {

  # Fill colours ------------------------------------------------------------
  if (is.null(fill_colors)) {
    n <- if (!is.null(color_var)) length(unique(df[[color_var]])) else NULL
    fill_colors <- slr_colors(n)
  }

  # If y_var != NULL, no summarise is needed. -------------------------------

  show_legend <- TRUE

  if (is.character(y_var)) {
    names(df)[names(df) == y_var] <- 'y'
    df$y2 <- 1

    if (!is.character(color_var)) {
      color_var <- "color_var"
      show_legend <- FALSE
    }

  } else{
  # Only one fill variabel used means no legend needed  ---------------------

    if (!is.character(color_var)) {
      color_var <- "color_var"
      show_legend <- FALSE
      df <-
        df %>%
        dplyr::group_by(.data[[x_var]]) %>%
        dplyr::summarise(y = dplyr::n()) %>%
        dplyr::mutate(y2 = sum(.data$y))
    } else{
  # Data transformations ----------------------------------------------------

      if (group_by_x_var) {
        df <-
          df %>%
          dplyr::group_by(.data[[x_var]], .data[[color_var]]) %>%
          dplyr::summarise(y = dplyr::n()) %>%
          dplyr::group_by(.data[[x_var]]) %>%
          dplyr::mutate(y2 = sum(.data$y))

      } else{
        df <-
          df %>%
          dplyr::group_by(.data[[x_var]], .data[[color_var]]) %>%
          dplyr::summarise(y = dplyr::n()) %>%
          dplyr::group_by(.data[[color_var]]) %>%
          dplyr::mutate(y2 = sum(.data$y))
      }
    }
  }

  # Ggplot ------------------------------------------------------------------

  lines <-
    ggplot2::ggplot(data = df) +
    ggplot2::scale_color_manual(
      values = fill_colors,
      labels = legend_labels,
      breaks = label_breaks,
      guide = ggplot2::guide_legend(nrow = legend_row, ncol = legend_col)
    )  +
    ggplot2::ylab(y_lab) +
    ggplot2::xlab(x_lab) +
    ggplot2::ggtitle(title, subtitle = subtitle) +
    theme_slr(subtitle = !is.null(subtitle),
              x_lab_exists = !is.null(x_lab),
              ...)

  if (y_percent) {
    y_breaks <- y_breaks / 100

    if (is.vector(y_lim)) {
      y_lim <- y_lim / 100
    }

    lines <-
      lines + ggplot2::geom_line(
        mapping = ggplot2::aes(
          x = .data[[x_var]],
          y = .data$y / .data$y2,
          color = if (utils::hasName(lines$data, color_var))
          .data[[color_var]] else color_var,
          group = if (utils::hasName(lines$data, color_var))
          .data[[color_var]] else color_var
        ),
        show.legend = show_legend,
        size = line_size
      ) +
      ggplot2::scale_y_continuous(
        labels = scales::percent_format(accuracy = percent_accuracy),
        breaks = seq(0, 1, by = y_breaks),
        limits = y_lim,
        expand = if(expand) waiver() else c(0,0)
      )

  } else{
    lines <-
      lines + ggplot2::geom_line(
        mapping = ggplot2::aes(
          x = .data[[x_var]],
          y = .data$y,
          color = if (utils::hasName(lines$data, color_var))
            .data[[color_var]] else color_var,
          group = if (utils::hasName(lines$data, color_var))
          .data[[color_var]] else color_var
        ),
        show.legend = show_legend,
        size = line_size
      ) +
      ggplot2::scale_y_continuous(
        breaks = seq(0, y_breaks_end, by = y_breaks),
        limits = y_lim,
        expand = if(expand) waiver() else c(0,0)
      )
  }
  lines
}