Functions … Putting it all together

Why functions?

• DRY principle: Don’t Repeat Yourself
  • It’s easier to fix/update code in one place than in 10
  • Lazy programming = best programming
• Use parameters to tweak function behavior (more customization)

Steps to make a function

• Write code to do the task once

• Modify code to make it more abstract

  • replace variable names with understandable parameter names
  • consider variations on the task and decide if you want to include them

• Wrap it in a function – define the function name and parameters

• Add error handling - check inputs, necessary conditions

Example: Solve the Quadratic Formula

Write some code

# Assume we have an equation ax^2 + bx + c = 0
# Solve for x

a = 1
b = 5
c = 2

x1 = (-b + sqrt(b^2-4*a*c))/(2*a)
x2 = (-b - sqrt(b^2-4*a*c))/(2*a)

c(x1, x2)

[1] -0.4384472 -4.5615528

Example: Solve the Quadratic Formula

Abstract the code

# Assume we have an equation ax^2 + bx + c = 0
# Solve for x

# Defined outside -- make more readable names
x2coeff = 1
xcoeff = 5
intcoeff = 2

## Extra parameter
under_sqrt = xcoeff^2-4*x2coeff*intcoeff
# need to check if under_sqrt is nonnegative
if (under_sqrt < 0) {
  print("Determinant is negative. Zeros will be imaginary.")
}

x1 = (-xcoeff + sqrt(under_sqrt))/(2*x2coeff)
x2 = (-xcoeff - sqrt(under_sqrt))/(2*x2coeff)

c(x1, x2)

[1] -0.4384472 -4.5615528

Example: Solve the Quadratic Formula

Wrap in a function

# Assume we have an equation ax^2 + bx + c = 0
# Solve for x
quad_formula <- function(x2coeff, xcoeff, intcoeff) {
  ## Extra parameter
  under_sqrt = xcoeff^2-4*x2coeff*intcoeff
  
  # need to check if under_sqrt is nonnegative
  if (under_sqrt < 0) {
    print("Determinant is negative. Zeros will be imaginary.")
  }
  
  x1 = (-xcoeff + sqrt(under_sqrt))/(2*x2coeff)
  x2 = (-xcoeff - sqrt(under_sqrt))/(2*x2coeff)
  
  return(c(x1, x2))
}

# Test the function
quad_formula(1, 5, 2)

[1] -0.4384472 -4.5615528

quad_formula(x2coeff = 1, xcoeff = 5, intcoeff = 2)

[1] -0.4384472 -4.5615528

Example: Solve the Quadratic Formula

Error Handling

# Assume we have an equation ax^2 + bx + c = 0
# Solve for x
quad_formula <- function(x2coeff, xcoeff, intcoeff) {
  # Ensure parameters are numeric before doing math with them
  stopifnot(is.numeric(x2coeff) & is.numeric(xcoeff) & is.numeric(intcoeff))
  
  ## Extra parameter
  under_sqrt = xcoeff^2-4*x2coeff*intcoeff
  
  # need to check if under_sqrt is nonnegative
  if (under_sqrt < 0) {
    print("Determinant is negative. Zeros will be imaginary.")
  }
  
  x1 = (-xcoeff + sqrt(under_sqrt))/(2*x2coeff)
  x2 = (-xcoeff - sqrt(under_sqrt))/(2*x2coeff)
  
  return(c(x1, x2))
}

# Test the function
quad_formula(1, 5, 2)

[1] -0.4384472 -4.5615528

quad_formula(x2coeff = 1, xcoeff = 5, intcoeff = 2)

[1] -0.4384472 -4.5615528

quad_formula("A", "B", "C")

Error in quad_formula("A", "B", "C"): is.numeric(x2coeff) & is.numeric(xcoeff) & is.numeric(intcoeff) is not TRUE