library(tidyverse)

#load in the data, keep the columns we want, and clean the data
load.subj.data <- function(prefix, subject, match_response, use_exp3=FALSE) {
  subj.data = read.subj.csv(sprintf("%s%02d_Test.csv", prefix, subject))
  
  begin_rows <- filter(subj.data, trigger == "BEGIN_TRIAL")
  end_rows <- filter(subj.data, trigger == "END_TRIAL")
  begin_rows$response <- end_rows$word
  subj.data <- select(begin_rows, trial, word, color, response)
  tStart <- strptime(as.character(begin_rows$time), "%Y.%m.%d-%H.%M.%OS")
  tStop <- strptime(as.character(end_rows$time), "%Y.%m.%d-%H.%M.%OS")
  
  options(digits.secs = 3)
  subj.data$rt <- as.numeric(tStop - tStart)
  subj.data$match <- subj.data$word == subj.data$color
  
  if (match_response == "RIGHT") {
    nonmatch_response = "LEFT"
  } else {
    nonmatch_response = "RIGHT"
  }
  subj.data$correct_response <- match_response
  subj.data$correct_response[subj.data$word != subj.data$color] <- nonmatch_response
  
  # Make a variable for correct/incorrect trial
  subj.data$correct <- (subj.data$match == 1 & subj.data$response == match_response) | (subj.data$match == 0 & subj.data$response == nonmatch_response)
  
  if (!use_exp3) {
    subj.data$hand <- "LEFT"
    subj.data$hand[subj.data$word >= 3] = "RIGHT"
    subj.data$hand <- as.factor(subj.data$hand)
  } else {
    subj.data$space <- "NONE"
    subj.data$space[subj.data$word == 0 | subj.data$word == 1] <- "LEFT"
    subj.data$space[subj.data$word == 2 | subj.data$word == 3] <- "BOTH"
    subj.data$space[subj.data$word == 4 | subj.data$word == 5] <- "RIGHT"
    subj.data$space <- as.factor(subj.data$space)
    
    subj.data$hand <- "NONE"
    subj.data$hand[subj.data$word == 0 | subj.data$word == 4 | subj.data$word == 2] = "LEFT"
    subj.data$hand[subj.data$word == 3 | subj.data$word == 1 | subj.data$word == 5] = "RIGHT"
    subj.data$hand <- as.factor(subj.data$hand)
  }
  
  subj.data$word <- as.factor(subj.data$word)
  subj.data$color <- as.factor(subj.data$color)
  subj.data$subject <- as.factor(subject)
  
  return(subj.data)
}

read.subj.csv <- function(filename) {
  subj.data <- read.csv(filename, stringsAsFactors = FALSE, header = FALSE)
  subj.data <- subj.data[,c(1,2,4,5,6,7)]
  colnames(subj.data) <- c("time", "line", "trigger", "trial", "word", "color")
  return(subj.data)
}

preprocess.data <- function(exp.data) {
  z <- exp.data$rt
  for (subject in unique(exp.data$subject)) {
    index <- exp.data$subject == subject
    z[index] <- (exp.data[index,]$rt - mean(exp.data[index,]$rt)) / sd(exp.data[index,]$rt)
  }
  exp.data <- subset(exp.data, abs(z) < 2.5)
  exp.data <- subset(exp.data, match == TRUE)
  exp.data <- subset(exp.data, correct == TRUE)
  exp.data$word <- as.numeric(exp.data$word)
  exp.data$color <- as.numeric(exp.data$color)
  return(exp.data)
}

# ------------------------------------------------------------------------------------------------------------
# Experiment 1

# Load in all subjects and group them together
s014 <- load.subj.data("Exp1/S", 14, "RIGHT")
s015 <- load.subj.data("Exp1/S", 15, "RIGHT")
s016 <- load.subj.data("Exp1/S", 16, "RIGHT")
s017 <- load.subj.data("Exp1/S", 17, "RIGHT")
s018 <- load.subj.data("Exp1/S", 18, "RIGHT")
s019 <- load.subj.data("Exp1/S", 19, "LEFT")
s020 <- load.subj.data("Exp1/S", 20, "LEFT")
s021 <- load.subj.data("Exp1/S", 21, "LEFT")
s022 <- load.subj.data("Exp1/S", 22, "LEFT")
s023 <- load.subj.data("Exp1/S", 23, "LEFT")
s024 <- load.subj.data("Exp1/S", 24, "RIGHT")
s025 <- load.subj.data("Exp1/S", 25, "RIGHT")
s026 <- load.subj.data("Exp1/S", 26, "RIGHT")
s027 <- load.subj.data("Exp1/S", 27, "RIGHT")
s028 <- load.subj.data("Exp1/S", 28, "RIGHT")
s029 <- load.subj.data("Exp1/S", 29, "LEFT")
s030 <- load.subj.data("Exp1/S", 30, "LEFT")
s031 <- load.subj.data("Exp1/S", 31, "LEFT")
s032 <- load.subj.data("Exp1/S", 32, "LEFT")
s033 <- load.subj.data("Exp1/S", 33, "LEFT")
s034 <- load.subj.data("Exp1/S", 34, "RIGHT")
s035 <- load.subj.data("Exp1/S", 35, "RIGHT")
s036 <- load.subj.data("Exp1/S", 36, "RIGHT")
s037 <- load.subj.data("Exp1/S", 37, "RIGHT")
s038 <- load.subj.data("Exp1/S", 38, "RIGHT")
s039 <- load.subj.data("Exp1/S", 39, "LEFT")
s040 <- load.subj.data("Exp1/S", 40, "LEFT")
exp1 <- rbind(s014, s015, s016, s017, s018, s019, s020, s021, s022, s023, s024, s025, s026, s027,
              s028, s029, s030, s031, s032, s033, s034, s035, s036, s037, s038, s039, s040)

# Remove outliers and only keep match trials and correct trials
exp1.pre <- preprocess.data(exp1)

write.csv(exp1.pre, "experiment1.csv", row.names=FALSE)

# ------------------------------------------------------------------------------------------------------------
# Experiment 2

# Load in subjects
t01 <- load.subj.data("Exp2/T", 1, "RIGHT")
t02 <- load.subj.data("Exp2/T", 2, "RIGHT")
t03 <- load.subj.data("Exp2/T", 3, "RIGHT")
t05 <- load.subj.data("Exp2/T", 5, "LEFT")
t07 <- load.subj.data("Exp2/T", 7, "LEFT")
t08 <- load.subj.data("Exp2/T", 8, "LEFT")
t10 <- load.subj.data("Exp2/T", 10, "RIGHT")
t11 <- load.subj.data("Exp2/T", 11, "RIGHT")
t12 <- load.subj.data("Exp2/T", 12, "RIGHT")
t13 <- load.subj.data("Exp2/T", 13, "RIGHT")
t14 <- load.subj.data("Exp2/T", 14, "RIGHT")
t15 <- load.subj.data("Exp2/T", 15, "LEFT")
t16 <- load.subj.data("Exp2/T", 16, "LEFT")
t17 <- load.subj.data("Exp2/T", 17, "LEFT")
t18 <- load.subj.data("Exp2/T", 18, "LEFT")
t19 <- load.subj.data("Exp2/T", 19, "LEFT")
t20 <- load.subj.data("Exp2/T", 20, "LEFT")
t21 <- load.subj.data("Exp2/T", 21, "RIGHT")
t23 <- load.subj.data("Exp2/T", 23, "RIGHT")
t24 <- load.subj.data("Exp2/T", 24, "LEFT")
t25 <- load.subj.data("Exp2/T", 25, "RIGHT")
t26 <- load.subj.data("Exp2/T", 26, "LEFT")
t27 <- load.subj.data("Exp2/T", 27, "LEFT")
t28 <- load.subj.data("Exp2/T", 28, "LEFT")
t29 <- load.subj.data("Exp2/T", 29, "LEFT")
t30 <- load.subj.data("Exp2/T", 30, "LEFT")
exp2 <- rbind(t01, t02, t03, t05, t07, t08, t10, t11, t12, t13, t14, t15, t16,
              t17, t18, t19, t20, t21, t23, t24, t25, t26, t27, t28, t29, t30)

exp2.pre <- preprocess.data(exp2)

write.csv(exp2.pre, "experiment2.csv", row.names=FALSE)

# ------------------------------------------------------------------------------------------------------------
# Experiment 2

# Load in subjects
s06 <- load.subj.data("Exp3/SVM", 6, "RIGHT", TRUE)
s07 <- load.subj.data("Exp3/SVM", 7, "RIGHT", TRUE)
s08 <- load.subj.data("Exp3/SVM", 8, "RIGHT", TRUE)
s09 <- load.subj.data("Exp3/SVM", 9, "RIGHT", TRUE)
s10 <- load.subj.data("Exp3/SVM", 10, "RIGHT", TRUE)
s11 <- load.subj.data("Exp3/SVM", 11, "LEFT", TRUE)
s12 <- load.subj.data("Exp3/SVM", 12, "LEFT", TRUE)
s13 <- load.subj.data("Exp3/SVM", 13, "LEFT", TRUE)
s14 <- load.subj.data("Exp3/SVM", 14, "LEFT", TRUE)
s15 <- load.subj.data("Exp3/SVM", 15, "LEFT", TRUE)
s16 <- load.subj.data("Exp3/SVM", 16, "RIGHT", TRUE)
s17 <- load.subj.data("Exp3/SVM", 17, "RIGHT", TRUE)
s18 <- load.subj.data("Exp3/SVM", 18, "RIGHT", TRUE)
s19 <- load.subj.data("Exp3/SVM", 19, "RIGHT", TRUE)
s20 <- load.subj.data("Exp3/SVM", 20, "LEFT", TRUE)
s21 <- load.subj.data("Exp3/SVM", 21, "RIGHT", TRUE)
s22 <- load.subj.data("Exp3/SVM", 22, "LEFT", TRUE)
s24 <- load.subj.data("Exp3/SVM", 24, "LEFT", TRUE)
s25 <- load.subj.data("Exp3/SVM", 25, "LEFT", TRUE)
s26 <- load.subj.data("Exp3/SVM", 26, "RIGHT", TRUE)
s27 <- load.subj.data("Exp3/SVM", 27, "RIGHT", TRUE)
s28 <- load.subj.data("Exp3/SVM", 28, "RIGHT", TRUE)
s29 <- load.subj.data("Exp3/SVM", 29, "RIGHT", TRUE)
s30 <- load.subj.data("Exp3/SVM", 30, "RIGHT", TRUE)
s31 <- load.subj.data("Exp3/SVM", 31, "LEFT", TRUE)
s32 <- load.subj.data("Exp3/SVM", 32, "LEFT", TRUE)
s33 <- load.subj.data("Exp3/SVM", 33, "LEFT", TRUE)
s34 <- load.subj.data("Exp3/SVM", 34, "LEFT", TRUE)
s35 <- load.subj.data("Exp3/SVM", 35, "LEFT", TRUE)
s36 <- load.subj.data("Exp3/SVM", 36, "LEFT", TRUE)
s37 <- load.subj.data("Exp3/SVM", 37, "RIGHT", TRUE)
s38 <- load.subj.data("Exp3/SVM", 38, "LEFT", TRUE)
s39 <- load.subj.data("Exp3/SVM", 39, "RIGHT", TRUE)
s40 <- load.subj.data("Exp3/SVM", 40, "RIGHT", TRUE)
s42 <- load.subj.data("Exp3/SVM", 42, "LEFT", TRUE)
s43 <- load.subj.data("Exp3/SVM", 43, "LEFT", TRUE)
s44 <- load.subj.data("Exp3/SVM", 44, "LEFT", TRUE)
s45 <- load.subj.data("Exp3/SVM", 45, "LEFT", TRUE)
s46 <- load.subj.data("Exp3/SVM", 46, "LEFT", TRUE)
s47 <- load.subj.data("Exp3/SVM", 47, "RIGHT", TRUE)
s48 <- load.subj.data("Exp3/SVM", 48, "RIGHT", TRUE)
s49 <- load.subj.data("Exp3/SVM", 49, "RIGHT", TRUE)
s50 <- load.subj.data("Exp3/SVM", 50, "RIGHT", TRUE)


exp3 <- rbind(s06, s07, s08, s09, s10, s11, s12, s13, s14, s15, s16, s17, s18, s19, s20, s21, s22,
              s24, s25, s26, s27, s28, s29, s30, s31, s32, s33, s34, s35, s36, s37, s38, s39, s40, s42,
              s43, s44, s45, s46, s47, s48, s49, s50)

exp3.pre <- preprocess.data(exp3)

write.csv(exp3.pre, "experiment3.csv", row.names=FALSE)