inst/create_linefigures_individual_GRINTA.R

In uashogeschoolutrecht/citrulliner: This package holds the data, the figures and R code for the 'Citrulline' paper

#' Create a summary line graph from a single analyte
#' 
#' A function to create a summary line graph from a 
#' dataset based on a single analyte. 
#' 
#' @param DF A data.frame containing data for a single analyte
#' @param palette_graph A palette with 5 colours
#'    
#' @return a ggplot2 object and the function writes to 
#' disk in the directory defined by the function 
#' saveInImageDirectory
#' 
#' @export 

# DF = DF_grinta
# palette_graph = grinta_palette
# ymax = ymax
# ymin = ymin

line_graph_indiv_GRINTA <- function(DF, 
                                    palette_graph, 
                                    ymax, 
                                    ymin){

  ##############################
  # load library & functions
  ##############################
  library(RColorBrewer)
  library(wesanderson)
  library(tidyverse)
  library(xlsx)
  library(dplyr)
  source(paste0(root, "/code/article_graphs/theme_individual.R"))
  source(file = file.path(root, 
                          "code", 
                          "article_graphs", 
                          "save_in_image_folder.R"))
  
  
  ################################
  # get analyte name
  ################################
  nameLine <- DF$analyte[1] %>%
    tolower()
  
  center <- DF$center[1]
  
  study <- paste0(toupper(as.character(DF$study[1])), "!") 
  
  nameFile <- paste(center, nameLine, study, sep = "_")
  
  
  ################################
  # get analyte metadata
  ################################  
  analyte_annotations <- readxl::read_excel(path = 
              file.path(
              root,
              "data-raw",
              "D016",
              "Copy of analytes_complete_ref_unit_SKa.xlsx"), 
              sheet = 1)
 
  y_axis_label <- analyte_annotations %>%
    dplyr::filter(tolower(analyte_short) == nameLine) %>%
    dplyr::select(units) %>%
    as.character()
  
  ###########################
  # long name for analyte 
  ###########################
  long_name <- analyte_annotations %>%
    dplyr::filter(tolower(analyte_short) == nameLine) %>%
    dplyr::select(analyte_long_name) %>%
    as.character()
 
   firstup <- function(x) {
    substr(x, 1, 1) <- toupper(substr(x, 1, 1))
    x
   }
 
   plot_title <- firstup(paste(long_name, study, center)) 
   #############################
   # axis limits
   #############################
   
   
   max_sem <- max(DF$sem)
   max_mean <- max(DF$mean_conc)
   
  # y_lim_max <- 1.2*(max_mean + max_sem)
  # y_lim_min <- -0.2*(max_mean)
  
  # y_lim_max <- ymax
  # y_lim_min <- ymin
   
    #############################
   # getting p-value from statistical analysis results (Eric Schoen)
   #############################
    #  statistics <- readxl::read_excel(path = 
     #                                 file.path(
      #                                  root,
      #                                  "data-raw",
      #                                  "D017_2",
      #                                  "FDR corrected P values.xlsx"), 
      #                              sheet = 1)
   
 #  p_value <- statistics %>%
  #   filter(analyte == long_name) %>%
   #  select(`FDR corrected P`) %>%
    # as.numeric() %>%
    # round(digits = 4)
    
  # p_value_annotations <- paste("P-value mixed model =", p_value)
   
  # DF$protocol <- as.factor(DF$protocol)
  # DF$time <- as.factor(DF$time)
  # DF$matrix <- as.factor(DF$matrix)
  
# palette_manual <- wes_palette(n=5, 
#                         name="Cavalcanti")
  
   plot <- ggplot(data = DF, aes(x = time, 
                               y = mean_conc,
                               colour = protocol)) +  
  
   geom_errorbar(aes(ymin = mean_conc - sem, 
                      ymax = mean_conc + sem),
                      width = 0.6,
                      position = position_dodge(width = 0.2),
                show.legend = FALSE) +
    
  geom_point(aes(shape = protocol), 
               size = 1.6, 
               position = position_dodge(width = 0.2),
             show.legend = FALSE) + 
  
  geom_line(aes(group = protocol), 
                size = 1.6, 
                position = position_dodge(width = 0.2)) +
    
  scale_colour_manual(labels = c("Rest", 
                                 "70%",
                                 "70% DH",
                                 "50%",
                                 "Interval"),
                     values = palette_graph)   +
   
  ## scale_color_identity(guide = "none") +
      
      #scale_colour_grey(start = 0, end = .85) +
     # facet_wrap(matrix ~ analyte, scales = "fixed", ncol = ncol) +
    
  scale_shape_manual(values=c(5,6, 15:17)) +
  scale_y_continuous(limits = c(ymin, ymax)) +
       # ggtitle(nameLine)  +
   
  theme_individual() +
  
   xlab("Time (hours)") + 
      
      ylab(paste("Concentration", paste0("(", 
                                       y_axis_label,
                                       ")"))) +
      
   #  geom_segment(aes(x=1, xend=1, y=-2, yend=-1.5), 
  #                arrow = arrow(length = unit(1, "cm"))) +
  
     
   #  annotate("segment", 
    #           x = 3, y = 0.25*y_lim_min, 
    #           xend = 3, yend = y_lim_min, 
    #           size = 0.7,
    #           arrow=arrow(length=unit(0.3, "cm"))) +
     
 #    annotate("text", 
#              x = 2, 
 #             y = 0.9*y_lim_max, 
  #            label = p_value_annotations) +
     
     
     ggtitle(plot_title)
    
  ##assigning a name to the file
 png <- paste(nameFile,"line.png", sep = "") 
#  eps <- paste(nameFile,"line.eps", sep = "") 
  svg <- paste(nameFile,"line.svg", sep = "") 

  ## saving the file to the imageDir
 saveInImageDirectory(svg)
 # saveInImageDirectory(eps)
  saveInImageDirectory(png)
  
  return(plot)
    
     
 }

# line_graph_indiv(DF)

#line_graph_indiv(DF, im_dir = im_dir, width = 5, dpi = 300, height = 5)

# line_graph(DF = subset_1, ncol = 4)

#    labs(title = paste(nameLine, "\n", sep=""), 
#         line = +4, 
#         units = "pt") + 
# geom_errorbar(aes(ymin = mean_conc - sem, 
#                  ymax = mean_conc + sem), 
#               width = .1, size = 0.5) +
#        scale_shape_manual(values=c(5,1,2,7,6)) +  

# apply self-made theme
# Change shapes
#    scale_linetype_manual(values=c("solid", 
#           "dotted", "longdash", "dotdash", "dashed"))