#
# Unit 8 - Data frames 
#
# Iterative production of a data frame 
#

# BACK TO REAL PROGRAMMING; SPECIFICALLY: LOOPS. 

# TAKE YOUR TIME, THIS IS IMPORTANT! 

# Loading data 
setwd("c:/R")
employees <- read.csv("employees.csv")
print(employees)

# One preliminary question 

# Question
# What does this function do: 
myMean <- function(vec) {
  mySum <- 0 
  for (val in vec) {
    mySum <- mySum + val
  }
  return (mySum / length(vec))
}

print(myMean(employees$SALARY))



# Problem:
# Build a new data frame in the following steps: 
# 1
# Produce 10 samples of observations with revision 
# on the basis of the data frame employees. 
# 2 
# For each sample. 
#   2.1 
#   calculate the average
#   of rows EDUC and YEAR. 
#   2.2. 
#   produce a vector containing the 
#   results of these two calculations
#   2.3 
#   Add the vector to the end of the 
#   constructed data frame. 


# Producing a new empty data frame 

# 1 - using data.frame 
newDF <- data.frame() 
print(newDF)

# 2 - with matrix()
# We will begin with this. 
newDF <- data.frame(matrix(NA, ncol=2, nrow = 3)  )
print(newDF)
print(class(newDF))

print(class(matrix(NA, ncol=2, nrow = 3)))

# Solving the problem 
newDF <- data.frame(matrix(NA, ncol=2, nrow = 10)  )
print(newDF)
# Problem:
# Build a new data frame in the following steps: 
# 1
# Produce 10 samples of observations with revision 
# on the basis of the data frame employees. 
for (iteration in 1:10) {
  # Producing sample of observations 
  inds <- sample(1:nrow(employees), size = nrow(employees), replace = TRUE)
  mySample <- employees[inds,]
  # means of EDUC and SALARY 
  meanEDUC <- myMean(mySample$EDUC)
  meanSALARY <- myMean(mySample$SALARY)
  # producing a vector with both means 
  results <- c(meanEDUC, meanSALARY)
  newDF[iteration,] <- results
}
print(newDF)

employees[2, ] <- c(3547, "Yasmin", "Lee", 2, 13, 15555)
print(employees)

employees[11, ] <- c(3547, "Yasmin", "Lee", 2, 13, 15555)
print(employees)


# How would we do the above using the 
# first method for producing an empty data frame? 
newDF <- data.frame() 
print(newDF)
for (iteration in 1:10) {
  # Producing sample of observations 
  inds <- sample(1:nrow(employees), size = nrow(employees), replace = TRUE)
  mySample <- employees[inds,]
  # means of EDUC and SALARY 
  meanEDUC <- myMean(mySample$EDUC)
  meanSALARY <- myMean(mySample$SALARY)
  # producing a vector with both means 
  results <- c(meanEDUC, meanSALARY)
  newDF <- rbind(newDF, results)
}
print(data.frame(newDF))

vec1 <- c(3, 4, 5)
vec2 <- c(7, 8, 9)
myDF <- data.frame(rbind(vec1, vec2))
print(myDF)
vec3 <- c(8, 8, 9)
myDF <- rbind(myDF, vec3)
print(myDF)

myDF2 <- data.frame() 
myDF2 <- rbind(myDF2, vec3)
myDF2 <- rbind(myDF2, vec3)
print(myDF2)