R/create_plot_data.R

In till-tietz/rbow: Bag of Words and Document Frequency Analysis for Multiple Dictionaries

#'re-format output of bow_ci function to plottable format
#'
#'@name create_plot_data
#'
#'@param bstrap_output output of bow_ci function
#'@param phenomena character vector specifying which phenomena you wish to plot
#'@param descriptors character vector specifying which descriptors you wish to plot
#'@param meta_data a data frame of meta data for each analyzed text i.e year, author etc.
#'@param combine if TRUE combines plotting data for each text into one data frame; if false returns a data frame for each text; default FALSE
#'@return data frame or list of data frames
#'@export

create_plot_data <-
  function(bstrap_output,
           phenomena = NULL,
           descriptors = NULL,
           meta_data = NULL,
           combine = FALSE) {
    each_text <- function(x) {
      bstrap_output_i <- bstrap_output[[x]]
      n_phenomena <- unique(bstrap_output_i[, "phenomenon"])
      n_descriptors <- unique(bstrap_output_i[, "descriptor"])

      if(!is.null(phenomena)){
        bstrap_output_i <- bstrap_output_i[which(bstrap_output_i[,"phenomenon"] %in% phenomena),]
        n_phenomena <- n_phenomena[which(n_phenomena %in% phenomena)]
      }

      if(!is.null(descriptors)){
        bstrap_output_i <- bstrap_output_i[which(bstrap_output_i[,"descriptor"] %in% descriptors),]
        n_descriptors <- n_descriptors[which(n_descriptors %in% descriptors)]
      }

      if (length(n_phenomena) %% 2 == 0) {
        step_size <- (length(n_phenomena) / 2) / 10
        steps <-
          seq(
            from = 1 - step_size,
            to = 1 + step_size,
            by = 0.1
          )
        steps <- steps[!steps %in% 1]
      } else {
        step_size <- ((length(n_phenomena) - 1) / 2) / 10
        steps <-
          seq(
            from = 1 - step_size,
            to = 1 + step_size,
            by = 0.1
          )
      }

      steps <- rep(steps, each = length(n_descriptors))
      pseudo <-
        rep(c(0:(length(n_descriptors) - 1)), length(n_phenomena))
      pseudo_y <- steps + pseudo

      out <- bstrap_output_i
      out[, "pseudo_y"] <- pseudo_y

      if (!is.null(meta_data)) {
        meta_df <-
          as.data.frame(sapply(meta_data[x, ], function(x)
            rep(x, nrow(out))))
        out <- cbind(out, meta_df)
      } else {
        out <- out
      }
      out[,"text"] <- rep(names(bstrap_output[x]), nrow(out))
      return(out)
    }

    plot_data <- purrr::map(1:length(bstrap_output), ~ each_text(.x))

    if (combine == TRUE) {
      plot_data <- do.call(rbind, plot_data)
    } else {
      plot_data <- plot_data
    }
    return(plot_data)
  }