Building Logic:
Operators, Conditions, and Functions

IOC-R Week 4

Plan

• Recap Week 3
• Operators
• Conditions
• Functions
• Let’s Practice !

Recap Week 3

Brief Recap Week 3

• Data frames: row-column structure

df <- data.frame(x = 1:3, y = c("a", "b", "c"))
df

  x y
1 1 a
2 2 b
3 3 c

• Lists: store anything

my_list <- list(x = 1:3, y = letters[5:1])
my_list

$x
[1] 1 2 3

$y
[1] "e" "d" "c" "b" "a"

What are the outputs for following codes?

df$x
df[[2]]
df[3, ]
df[, "y", drop = FALSE]
df[1, 2]

my_list$x
my_list[1]
my_list["y"]
my_list[[2]][3]

list is often used to store meta data

Motivation for Week 4 - Why learn operators, conditions, and functions? - Real-world examples (e.g., filtering data, automating repetitive tasks, decision-making in analysis).

Operators

Assignment and Arithmetic Operators

• Assignment operators: <-, ->, =

x <- 1
x

[1] 1

2 -> x # not recommanded
x

[1] 2

y = "abc" # not recommanded
y

[1] "abc"

mean(x = 1:3) # assign value to function’s parameter

[1] 2

• Arithmetic operators: +, -, *, /, ^ (exponentiation), %% (remainder), etc.

10^3

[1] 1000

10 %% 3 # modulus (remainder) of a division operation

[1] 1

x <- c(2, 8, 3)
x - 1

[1] 1 7 2

x - 1:3

[1] 1 6 0

y <- c(6, 4, 1)
x + y

[1]  8 12  4

x * y

[1] 12 32  3

x / y

[1] 0.3333333 2.0000000 3.0000000

Comparison Operators

Element-wise comparison: ==, !=, >, <, >=, <=, return logical results.

10 == 3

[1] FALSE

10 != 3

[1] TRUE

10 > 3

[1] TRUE

10 >= 3

[1] TRUE

10 < 3

[1] FALSE

10 <= 3

[1] FALSE

x <- c(2, 8, 3)
y <- c(6, 8, 1)
x == y

[1] FALSE  TRUE FALSE

x != y

[1]  TRUE FALSE  TRUE

x > y

[1] FALSE FALSE  TRUE

x <= y

[1]  TRUE  TRUE FALSE

What are the expected results?

• x == 8
• y > 1
• a is a vector of numbers 1 to 5, a[a > 3] returns?

Logical Operators

Element-wise comparison: NOT (!), AND (&), OR (|), etc., returns logical values.

!TRUE

[1] FALSE

!FALSE

[1] TRUE

To combine two conditions A and B:

A	B	A & B	A | B
TRUE	TRUE	TRUE	TRUE
TRUE	FALSE	FALSE	TRUE
FALSE	TRUE	FALSE	TRUE
FALSE	FALSE	FALSE	FALSE

What are the expected results?

• (3 > 1) & (7 < 5)
• (3 > 1) | (7 < 5)

(3 > 1) & (7 < 5)

[1] FALSE

(3 > 1) | (7 < 5)

[1] TRUE

The %in% Operator

We use %in% to check if left-side values are present in right-side, it returns logical values.

## check membership
1 %in% 1:3

[1] TRUE

1:3 %in% 1

[1]  TRUE FALSE FALSE

"a" %in% c("abc", "a")

[1] TRUE

Data frame df has two columns x and y, what is the result of following code?

"col3" %in% colnames(df)

"col3" %in% colnames(df)

[1] FALSE

Bonus: Operator Precedence

Simplified precedence order¹ (highest to lowest):

1. Arithmetic (^ > *, / > +, -)
2. Relational (>, <, >=, <=, ==, !=)
3. Membership (%in%)
4. Logical (! > & > |)

Order can be change using ():

1 + 2 * 3

[1] 7

(1 + 2) * 3

[1] 9

TRUE | TRUE & FALSE

[1] TRUE

(TRUE | TRUE) & FALSE

[1] FALSE

1. For more detail, check the operator documentation.

any(), all() and which()

Given a set of logical vectors:

-3:3

[1] -3 -2 -1  0  1  2  3

-3:3 > 0

[1] FALSE FALSE FALSE FALSE  TRUE  TRUE  TRUE

• any(): is at least one of the values TRUE?

any(-3:3 > 0)

[1] TRUE

• all(): are all of the values TRUE?

all(-3:3 > 0)

[1] FALSE

• which(): return indices of TRUE values.

which(-3:3 > 0)

[1] 5 6 7

We have a vector log2FoldChange <- c(1.2, -0.5, 0.9, 0.7, -1.1), what are the expected results for the following codes?

• which(log2FoldChange > 0.8)
• any(log2FoldChange > 0.8)
• all(log2FoldChange > 0)

Conditions

Conditional Statements

Conditional statements allow us to make decisions based on logical conditions, guiding how the code behaves in different scenarios.

flowchart LR
  A{temperature > 37} --> B(TRUE)
  A --> C(FALSE)
  B --> D[Fever]
  C --> E[Normal]

flowchart LR
  A{log2FC > 0} --> B(TRUE)
  A --> C(FALSE)
  B --> D[Up-regulated]
  C --> E[Not up-regulated]

if and if else

Syntax:

## if statement
if (condition) {
  # code to run if `condition` is TRUE
}

## if else statement
if (condition) {
  # code to run if `condition` is TRUE
} else {
  # code to run if `condition` is FALSE
}

• You can use if without else but never in the opposite way.
• The condition must evaluate to exactly one logical value (TRUE or FALSE) and cannot be NA.

flowchart LR
  A{log2FC > 0} --> B(TRUE)
  A --> C(FALSE)
  B --> D[Up-regulated]
  C --> E[Not up-regulated]

log2FC <- 2.5
if (log2FC > 0) {
  # code to run if condition is TRUE
  print("Up-regulated")
} else {
  # code to run if condition is FALSE
  print("Not up-regulated")
}

[1] "Up-regulated"

What will you get when log2FC is 0?

[1] "Not up-regulated"

The ifelse() Function

Syntax:

ifelse(test, yes, no)

Example:

log2FC <- -1
ifelse(test = log2FC > 0, yes = "Up-regulated", no = "Not up-regulated")

[1] "Not up-regulated"

# nested condition
ifelse(
  test = log2FC > 0,
  yes = "Up-regulated",
  no = ifelse(log2FC < 0, yes = "Down-regulated", no = "No change")
)

[1] "Down-regulated"

The “test” parameter can be a logical vector.

vec_lfc <- c(-1, 0, 1)
ifelse(vec_lfc > 0, yes = "Up-regulated", no = "Not up-regulated")

[1] "Not up-regulated" "Not up-regulated" "Up-regulated"    

Functions

What is a Function?

Functions = Reusable blocks of code.

Call a function: write function name followed by (), including any required arguments inside ().

• Built-in functions

mean(x = 1:3)

[1] 2

Functions that are part of R’s base installation, always available without needing to load any additional packages.

• Custom functions

add_one <- function(x) {x + 1}
add_one(x = 1:3)

[1] 2 3 4

• Functions from additional packages

library(readr)
read_csv(file = "path/to/data.csv")

We’ll talk about packages next week!

Use ? or help() to view the function’s documentation, e.g.: ?mean, help(mean)

Anatomy of a Function

Syntax:

function_name <- function(arguments) {
  # function body (code for tasks)
  return(result)
}

• The function name, be concise and preferably a verb.
• The reserved token function() and argument(s).
• The curly brackets {} frame the function body.
• The return() (usually) at the end will return a result.

What is the composition of the following function?

add_one <- function(x) {
  result <- x + 1
  return(result)
}

Function Parameters

A function can have 0, 1 or multiple parameters, with or without default values.

• No parameter: perform an action without needing input

say_hello <- function() {
  print("Hello!")
}
say_hello()

[1] "Hello!"

• 1 paramter:

add_one <- function(x = 5) {
  return(x + 1)
}
add_one() # use default value of x

[1] 6

add_one(x = 1:3) # use custom values for x

[1] 2 3 4

vec <- 1:2
add_one(x = vec) # use custom values from a variable

[1] 2 3

• Multiple parameters

divise <- function(numerator, denominator) {
  return(numerator/denominator)
}
# Provide argument values in the expected order
divise(numerator = 1, denominator = 2)

[1] 0.5

divise(1, 2)

[1] 0.5

# Provide argument values in a different order
divise(2, 1)

[1] 2

divise(denominator = 2, numerator = 1)

[1] 0.5

Function Body

Example: a function to calculate geometric mean¹.

geo_mean <- function(x) {
  # calculate geometric mean
  result <- exp(mean(log(x)))
  return(result)
}
geo_mean(c(1:10, 50))

[1] 5.633703

mean(c(1:10, 50))

[1] 9.545455

But, geometric mean can only be calculated for positive numbers.

Use a condition to ensure this.

geo_mean <- function(x) {
  # keep only positive values
  if (any(x <= 0)) {
    x <- x[x > 0]
  }
  # calculate geometric mean
  result <- exp(mean(log(x)))
  return(result)
}
geo_mean(c(1:10, 50, -1, 0))

[1] 5.633703

geom mean is more suitable for data with a log-normal distribution or when values have outliers or large differences in scale.

1. Compare to arithmetic mean, geometric mean is more suitable for data with outliers or large differences in scale.

Local vs. Global Variables

• Local variable: defined inside the function.
• Global variable: defined outside and accessible anywhere.

x <- 5
add_one <- function() {
  x <- 10
  return(x + 1)
}
add_one()  # Local, x is 10

[1] 11

x  # Global, x is 5

[1] 5

Scope in R, importance of proper scoping in functions.

Let’s Practice !

Today’s Goals

• Understand and use operators to filter data with precision.
• Understand function’s basic structure, learn to use new functions, write functions to automate tasks.