R/plot_rois_in_data.R

In matiasandina/nobrainr: What the Package Does (One Line, Title Case)

# This will plot what rois you have in the data
# it's similar to plot_roi_counts but it only checks for value != NA
#' @export
plot_rois_in_data <- function(count_df, AP, half = TRUE) {
  stopifnot(is.numeric(AP))
  stopifnot(is.data.frame(count_df))
  # generate the data using helper functions
  df <- generate_atlas_data(AP, half = half)
  # make sure we only have the plate we are looking for
  count_df <-
    dplyr::filter(count_df, mm_from_bregma == roundAP(AP)) %>%
    # select the columns we care about
    dplyr::select(value, parent, side)
  if (nrow(count_df) <= 0) {
    usethis::ui_stop("No counts found for AP={AP} in `count_df`")
  }
  # bind with counts
  df <- df %>%
    dplyr::left_join(count_df, by = c("parent", "side"))

  # construct the plot
  base_plot <- ggplot2::ggplot() +
    ggplot2::coord_equal() +
    ggplot2::theme_void()
  # user feedback
  usethis::ui_info("Coloring by `parent` in `count_df`")
  # add layers to the plot
  right_plot <-
    base_plot +
    ggplot2::geom_polygon(
      data = dplyr::filter(df,
                    parent != "root"),
      ggplot2::aes(x, y,
                   group = interaction(side, path)),
      fill = "gray90",
      color = "black"
    ) +
    ggplot2::geom_polygon(
      data = dplyr::filter(df,
                    parent != "root",!is.na(value)),
      ggplot2::aes(x, y,
                   group = interaction(side, path),
                   fill = parent),
      color = "black"
    )
  return(right_plot)

}