R/plot_TS.R

In ICRISAT-GEMS/SpaTemHTP: Processing and utilization of HTP data

###########
# TS_plot #
###########

#' Plot G-BLUEs time series data
#' 
#' Plot G-BLUEs time series data.
#' 
#' The function plots the times series data of phenotyping data per genotype
#' replicate or averaged over genotype replicate using the ggplot2 package.
#' 
#' @param data_TS \code{numeric} \code{matrix} of dimension N_genotype 
#' x N_days containing the phenotype information. the different columns
#' represent the different time points (e.g. days, hours) of measurement.
#' 
#' @param trend Optional \code{numeric} \code{matrix} of dimension n_days containing the
#' values of the trend to be ploted on the top of the time series Default = NULL.
#' 
#' @param h2 Optional heritability \code{numeric vector} that will be used to
#' indicate the day with the highest heritability. Default = NULL.
#' 
#' @param TW_lim Optional time window limits \code{numeric vector} that will be
#' used to indicate the time window limit days. Those values can be obtained with
#' \code{\link{cpa_getOTW_2}}. Default = NULL.
#' 
#' @param main \code{Character} string representing the title of the graph. Default =  'G-BLUEs TS'.
#'
#' @return
#' 
#' Plot of the G-BLUEs time series. The optional trend will be ploted in red.
#' The heritability will be ploted in dashed blue and time window limits in black.
#'
#' @author Soumyashree Kar, Vincent Garin
#' 
#' @seealso
#' 
#' \code{\link{SpaTemHTP_proc}}
#'
#' @examples
#'
#' data(SG_PH_data)
#' data <- SG_PH_data
#' geno_id <- data$genotype
#' 
#' g_means <- matrix(NA, 384, 23)
#' 
#' for(i in 1:23){
#'   
#'   g_av <- tapply(X = data[, i + 5], INDEX = geno_id,
#'                  FUN = function(x) mean(x, na.rm = TRUE))
#'   g_means[, i] <- g_av
#'   
#' }
#' 
#' rownames(g_means) <- names(g_av)
#' 
#' plot_TS(data_TS = g_means, main = 'Raw data genotype means')
#' 
#'
#' @export
#'


plot_TS <- function(data_TS, trend = NULL, h2 = NULL, TW_lim = NULL,
                    main = 'G-BLUEs TS'){
  
  if(!is.null(h2)){
    
    h2_lines <- geom_vline(xintercept = which.max(h2), linetype="dotted", 
                           color = "blue", size=1.5)
    
  } else {h2_lines <- NULL}
  
  if(!is.null(TW_lim)){
    
    TW_lines <- geom_vline(xintercept = TW_lim, color = "black", size=1)
    
  } else {TW_lines <- NULL}
  
  
  n_days <- dim(data_TS)[2]
  
  dt <- data.frame(geno = rep(rownames(data_TS), n_days),
                   day = rep(1:n_days, each = dim(data_TS)[1]),
                   trait = c(data_TS))
  
  if(is.null(trend)){
    
    plot <- ggplot(data = dt, aes(x = day, y = trait, group = geno)) +
      geom_line(aes(group = geno)) + ggtitle(main) + h2_lines + TW_lines
    
    print(plot)
    
  } else {
    
    
    dt_trend <- data.frame(trait = trend, day = 1:n_days, geno = 'G_av')
    
    plot <- ggplot(data = dt, aes(x = day, y = trait, group = geno)) +
      geom_line(aes(group = geno)) + geom_line(data = dt_trend, color = 'red', size = 1) +
      ggtitle(main) + h2_lines + TW_lines
    
    print(plot)
    
  }
  
  
}