Data Science Using R (Spring 2024) - Session 2

Session-02-Script.R

# This is the script that we developed during the session on March 21, 2024

# 1. Basic commands-----------
# This is an addition:
2 + 4
4 - 9 # This is substraction
4/9
3*9
2**3

# This is 
#  a comment that spans
#   over several lines.

# 2. Intermediate task I -----------
5+12
(2*3)**2
2*5.8 # Beware: decimal sign in R is '.', not ','
(8**2 + 5**4) / 3

## 3. Assignments  -----------
assign("int_result", 2 + 4)
int_result <- 2 + 4 # Does the same as previous line

a1 <- 2
a1 <- 4
a1 # Second assignment got overridden

# 4. Intermediate task II -----------
a_ <- 2+3
b_ <- (5*a_)/2
c_ <- (b_+1)**2
d_ <- sqrt(c_) # c_**0.5

# Hint: if you get the error "Object not found" such as here:
obj_x
# ...then you should always check the following:
#. 1. Did you misspell the name?
#. 2. Did you forget the do the assignment in the script? -> search for the name 
# Also, if you get a weird error, always check to which objects the involved 
#  names actually point

# 5. Functions -----------
t_vec <- c(1, 2, 3, 4)
# The following four lines do the same thing:
mean(t_vec)
mean(c(1,2,3,4))
mean(x = t_vec) # The name of the mandatory argument 'x' is used explicitly
mean(x = c(1,2,3,4))
help(mean) # Get information about arguments

# Using optional arguments:
t_vec <- c(1,2,3,4, NA)

mean(t_vec)
mean(t_vec, na.rm = FALSE)
mean(t_vec, na.rm = TRUE)
# From the fact that `mean(t_vec, na.rm = FALSE)` gives the same result as
#  `mean(t_vec)` we can conclude that by default `na.rm = FALSE`.

# 6. Intermediate task III -----------
task_vector <- c(-2, 2, 4, 6, 9, NA)

median(task_vector)
median(task_vector, na.rm = TRUE)
is.na(task_vector)
anyNA(task_vector)
sum(task_vector)
sum(task_vector, na.rm = TRUE)
# If in doubt use `?` or `help()`. 
# This also explains that var() computes the sample variance:
var(task_vector, na.rm = TRUE)
# To compute the population variance you can write your own function:
var_pop <- function(input_vector) {
  mean((input_vector - mean(input_vector))**2)
}
# Or correct the result of var():
n_obs <- length(task_vector) # get the number of observations
var(task_vector)*(n_obs-1)/n_obs
# Although this is slightly less accurate because we count the NA

# 7. Defining your own functions --------------
add_two_numbers <- function(nb1, nb2){
  result <- nb1 + nb2
  return(result)
}
add_two_numbers(1, 4)
add_two_numbers(nb1 = 1, nb2 = 4) # equivalent, but mandatory arguments set explicitly
nb1 # Not defined outside the function body!

# 8. Defining function - exercise I --------------
compute_eq <- function(x_1, x_2){
  result <- x_1**2 + 2*x_1*x_2 + x_2**2
  return(result)
}
compute_eq(2, 4)

# 8. Defining function - exercise II --------------
normalize_vec <- function(input_vector){
  (input_vector - min(input_vector)) / (max(input_vector) - min(input_vector))
}
normalize_vec(c(1,2,3,4))