R/mcfa_scatterplot.R

In RSDA: R to Symbolic Data Analysis

#' Plot Interval Scatterplot
#'
#' @param x symbolic table with only one column.
#' @param y symbolic table with only one column.
#' @param sym.data original symbolic table.
#' @param pos.var column number of the variables to be plotted.
#'
#' @examples
#' data("ex_mcfa1")
#' sym.table <- classic.to.sym(ex_mcfa1,
#'   concept = suspect,
#'   hair = sym.set(hair),
#'   eyes = sym.set(eyes),
#'   region = sym.set(region)
#' )
#'
#' res <- sym.mcfa(sym.table, c(1, 2))
#' mcfa.scatterplot(res[, 2], res[, 3], sym.data = sym.table, pos.var = c(1, 2))
#' @export
#' @importFrom  purrr flatten_chr map2
#' @importFrom ggplot2 ggplot geom_rect aes geom_text theme_minimal labs theme

mcfa.scatterplot <- function(x, y, sym.data, pos.var) {

  x <- to.v2(x)
  y <- to.v2(y)
  sym.data <- to.v2(sym.data)
  var.names <- c()
  n.vars <- c()
  for (i in pos.var) {
    var.names <- c(var.names, sym.data[, i]$sym.var.names)
    n.vars <- c(n.vars, ncol(sym.data[, i]$data))
  }

  df <- cbind(x$data, y$data)
  colnames(df) <- c("C1", "C1.1", "C2", "C2.1")
  df$var.name <- purrr::flatten_chr(purrr::map2(var.names, n.vars, ~ rep(.x, .y)))
  df$cat.name <- x$sym.obj.names

  ggplot(data = df) +
    geom_rect(mapping = aes(xmin = C1, xmax = C1.1, ymin = C2, ymax = C2.1, color = var.name), fill = NA, size = 1) +
    geom_text(aes(label = cat.name, x = C1, y = C2)) +
    theme_minimal() +
    labs(color = "Variables") +
    theme(legend.position = "bottom")
}