###########################################################
# LOAD DATA AND LIBRARIES
###########################################################

library(plyr)
library(tidyverse)
library(gridExtra)
library(data.table)
#install.packages("remotes")
#remotes::install_github("wilkelab/cowplot")

library(cowplot)

#source("CI-Functions.R")
source("CI-Functions-Bonferroni.R")

defaultpath <- dirname(rstudioapi::getActiveDocumentContext()$path)
setwd(defaultpath)

create_row <- function(data_base, current_task, measure_name, mycolor) {
  
  col_name <- paste0("mean_", measure_name, sep="")
  elements <- data_base %>% select(user, block, measure) %>% arrange(user, block)
  #View(elements)
  setnames(elements, old=c("measure"), new=c(col_name))
  elements <- spread(elements, key = block, value = col_name)
  setnames(elements, old=c("1","2", "3"), new=c("First","Second", "Third"))
  
  #########
  # stats #
  #########
  
  data <- elements
  
  if (mean (data$First) == 0 ) {
    techniqueA <- c(0.000000,0.000000,0.000000)
  } else {
    techniqueA <- bootstrapMeanCI(data$First)  
  }
  if (mean (data$Second) == 0 ) {
    techniqueB <- c(0.000000,0.000000,0.000000)
  } else {
    techniqueB <- bootstrapMeanCI(data$Second)  
  }
  if (mean (data$Third) == 0 ) {
    techniqueC <- c(0.000000,0.000000,0.000000)
  } else {
    techniqueC <- bootstrapMeanCI(data$Third)  
  }
  
  analysisData <- c()
  analysisData$name <- c("Third","Second","First") 
  analysisData$pointEstimate <- c(techniqueC[1], techniqueB[1], techniqueA[1])
  analysisData$ci.max <- c(techniqueC[3], techniqueB[3], techniqueA[3])
  analysisData$ci.min <- c(techniqueC[2], techniqueB[2], techniqueA[2])
  
  datatoprint <- data.frame(factor(analysisData$name),analysisData$pointEstimate, analysisData$ci.min, analysisData$ci.max)
  colnames(datatoprint) <- c("vis", col_name, "lowerBound_CI", "upperBound_CI ") #We use the name mean_error for the value of the mean even though it's not a error, it's just to parse the data for the plot
  
  path  = paste0("plots/order_session/")
  filename = paste0("order_", measure_name, "_means_", current_task, sep="")
  
  write.table(datatoprint, paste0(path, "printed_",filename,".txt",seq=""), sep=",",row.names=FALSE)
  mean_plot <- barChart(datatoprint, analysisData$name, nbTechs = 3, ymin = 0, ymax = 50, "", "", mycolor=mycolor)
  #ggsave(paste0(path,"plot_",filename,".pdf",seq=""), mean_plot, device = "pdf", width=5, height=2)
  
  # CIs with adapted alpha value for multiple comparisons not needed here
  # checks for situations where error rate is 0 everywhere, else Bootstrap
  if (mean(data$Second) == 0 & mean(data$First) == 0) {
    diffBA <- c(0.000000,0.000000,0.000000,8,0.000000,0.000000)
  } else {
    diffBA = bootstrapMeanCI_corr(data$Second - data$First, 1)
  }
  if (mean(data$Third) == 0 &  mean(data$Second) == 0) {
    diffCB <- c(0.000000,0.000000,0.000000,8,0.000000,0.000000)
  } else {
    diffCB = bootstrapMeanCI_corr(data$Third - data$Second, 1)
  }
  if (mean(data$Third) == 0 & mean(data$First)) {
    diffCA <- c(0.000000,0.000000,0.000000,8,0.000000,0.000000)
  } else {
    diffCA = bootstrapMeanCI_corr(data$Third - data$First, 1)
  }
  
  analysisData <- c()
  analysisData$name <- c("Third-First","Third-Second","Second-First")
  analysisData$pointEstimate <- c(diffCA[1], diffCB[1], diffBA[1])
  analysisData$ci.max <- c(diffCA[3], diffCB[3], diffBA[3])
  analysisData$ci.min <- c(diffCA[2], diffCB[2], diffBA[2])
  analysisData$level <- c(diffCA[4], diffCB[4], diffBA[4])
  analysisData$ci_corr.max <- c(diffCA[6], diffCB[6], diffBA[6])
  analysisData$ci_corr.min <- c(diffCA[5], diffCB[5], diffBA[5])
  
  
  datatoprint <- data.frame(factor(analysisData$name), analysisData$pointEstimate, analysisData$ci.max, analysisData$ci.min, analysisData$level, analysisData$ci_corr.max, analysisData$ci_corr.min)
  colnames(datatoprint) <- c("technique", col_name, "lowerBound_CI", "upperBound_CI", "corrected_CI", "lowerBound_CI_corr", "upperBound_CI_corr") #We use the name mean_error for the value of the mean even though it's not a error, it's just to parse the data for the plot
  
  path  = paste0("plots/order_session/")
  filename = paste0("order_", measure_name, "_diffs_",current_task)
  
  write.table(datatoprint, paste0(path,"printed_",filename,".txt",seq=""), sep=",",row.names=FALSE)
  
  diff_plot <- barChart(datatoprint, analysisData$name, nbTechs = 3, ymin = -20, ymax = 20, "", "", mycolor=mycolor)
  #ggsave(paste0(path,"plot_",filename,".pdf",seq=""), diff_plot, device = "pdf", width=5, height=2)
  
  row <- grid.arrange(mean_plot, diff_plot, ncol=2)
  return(row)
}

all_data <- read.table("user_answers_ms.csv", header=T, sep=",")
# Remove training
all_data <- all_data %>% filter(question_config != 'T') %>%
  mutate(vis=case_when(vis == 'Symbol' ~ "Gly", 
                       vis == 'Dorling' ~ "Dor", 
                       vis == 'Barchart' ~ "Bar"))
# Summarize by task
data_base <- all_data %>% select(user, vis, block, question_code, time, is_correct)
data_base <- data_base %>% group_by(user, question_code, block) %>% summarise(time=mean(time), error=sum(is_correct=='False')/3)
data_base <- data_base %>%  mutate(time=time/1000,
                                   error=error*100)


data_time <- data_base %>% group_by(user, block) %>% summarise(measure=mean(time))
row_time <- create_row(data_time, "all", 'time', 'dodgerblue2')

data_error <- data_base %>% group_by(user, block) %>% summarise(measure=mean(error))
row_error <-create_row(data_error, "all", 'time', 'darkorange3') #this uses error but not in the name, for some reason I got error

grid <- grid.arrange(row_time, row_error, ncol=1)

filename <- "plots/order_session/all_time_error.pdf"
ggsave(filename = filename, grid, device = "pdf", width=10, height=6)