Intro to Data Science

Lecture 2 – Packages, Structure, & Conditionals

A Guide to Your Process

Scheduling

Learning Objectives

Practice

Supporting Information

Class Discussion

Today’s Plan

• R Packages
• Class vs. Structure
• Using Vectors
• Conditionals

Today’s Learning Objectives

After today’s session you will be able to:

• Load and use an R package
• Define the difference(s) between object class and structure
• Create and manipulate vectors
• Write conditional statements
• Manage missing data in objects with conditionals

R Package Background

• R packages are groups of functions developed by users

• Packages have no defined depth or breadth requirements
  • A package could be a single, simple function
  • Or a complex ecosystem of inter-related functions

• Packages can be installed by any R user for free!

• R is versatile and powerful (in part) because of contributed packages

Package Locations

• There are two main homes for R packages

1. Comprehensive R Archive Network
   • A.k.a. “CRAN” | cran.rstudio.com

2. GitHub
   • github.com

CRAN vs. GitHub

CRAN

• Currently >20,000 packages live here

• Strict rules for packages to be allowed

• These are “official” packages

GitHub

CRAN vs. GitHub

CRAN

• Currently >20,000 packages live here

• Strict rules for packages to be allowed

• These are “official” packages

GitHub

• Unknown number of packages here (no centralized record retained)

• No mandatory quality control tests to be available here

• Packages usually work but they don’t have the same quality control as CRAN packages

Using Packages

• In order to use a package, you must:

1. Install the desired package
   • Done once per computer

2. Load the package into R
   • Done every time you re-open RStudio

Using Packages: Specific Steps

# Install desired package
install.packages("dplyr")

# Load that package
library(dplyr)

• install.packages requires the package name be in quotes

• library requires requires unquoted names.

Package Analogy

• install.packages = buying a set of tools from the store and putting them in your home

• library = moving the tools that you already own to your workbench

• You only buy the tools once but every time you start work you need to bring them back to your work area!

Practice: Packages

• Make a new script for this week’s lecture!
  • Save it in your RStudio Project folder for this course
  • Make sure it has “lecture” and “2” in the file name

• Install the palmerpenguins package
  • Remember to put quotes around the package name!

• Load the package with the library function
  • Once loaded, run ?palmerpenguins to see the package-level help file

Note on Function Names

• Functions are not required to have unique names across all packages

• Risk of using a different function than intended
  • Best case: creates an error and forces you to catch the mistake
  • Worst case: silently does something wrong

Function Namespacing

• Functions can be “namespaced” to specify which package the function comes from

• Namespacing guarantees you use the function from the desired package

• Done with two colons (:) between the package and function names
  • E.g., package::function()

Practice: Namespacing

• palmerpenguins includes an example dataset on penguins

• Run the following code
  • peng_df <- palmerpenguins::penguins

• Install and load the dplyr package
  • dplyr has a function called glimpse that shows you core structures of data

• Namespace glimpse and run it on peng_df

Temperature Check

How are you Feeling?

Class versus Structure

• Objects have both class and structure

• Class = the type of object it is
  • E.g., dataframe, integer, character, etc.

• Structure = the dimensions and “shape” of the data
  • E.g., Number of rows / columns, length, etc.

• Both class and structure affect what you can do with or to a given object!

Checking Class/Structure

• Best to check class and structure of an object to ensure functions will work

Check Class

Check Structure

Checking Class/Structure

• Best to check class and structure of an object to ensure functions will work

Check Class

• Use the class function

# Checking class of 'my_obj'
class(my_obj)

Check Structure

Checking Class/Structure

• Best to check class and structure of an object to ensure functions will work

Check Class

• Use the class function

# Checking class of 'my_obj'
class(my_obj)

Check Structure

• Use either the str function or dplyr::glimpse

# Checking structure of 'my_obj' (with base R)
str(my_obj)

# Checking structure of 'my_obj' (with `dplyr`)
dplyr::glimpse(my_obj)

Practice: Structure

• What is the class of ‘peng_df’?

• What is the structure?
  • What information is included when you check?

• What happens when you check the class of a function?
  • Run class(class)

• What happens when you check the structure of a function?
  • Run str(str)

Using Vectors: Coordinates

• Vector structure is expressed as “length”
  • Vector length = number of elements in the vector
  • Dataframe length = number of rows

• Bracket notation can be used to navigate vectors

# Make a vector
my_vec <- c("a", "b", "c", "d", "e")

# Use bracket notation to retrieve one element
my_vec[3]

[1] "c"

Using Vectors: Coordinates Cont.

• Bracket notation accepts vectors of coordinates

# Use bracket notation to retrieve several elements
my_vec[c(1, 3, 5)]

[1] "a" "c" "e"

• You can also grab the same element more than once!

# Use bracket notation to retrieve one element multiple times
my_vec[c(1, 1, 1)]

[1] "a" "a" "a"

Practice: Vector Navigation

• Base R has a built-in vector of letters called letters
  • Assign letters to an object called my_vec

• Check the length of my_vec using the length function

• Identify the 10^th element of my_vec using bracket notation

• Identify the 8^th, 5^th, 12^th, 12^th (again), and 15^th elements of my_vec
  • Use concatenation inside of the brackets (with c)!

Temperature Check

How are you Feeling?

Conditionals

• You can write code that runs only if an ‘if statement’ is true
  • Otherwise that chunk of code is skipped!

• This allows you to write flexible code that can handle any outcome that you can anticipate!
  • Particularly useful for subsetting data based on the contents of a column

• These ‘if statements’ are called conditionals

• The answer to a conditional must be either TRUE or FALSE

Fundamentals: EQUAL

• Are two things exactly equal?

# Check a conditional
"hello" == "hello"

[1] TRUE

• Uses == operator
  • Just two equal signs

Fundamentals: OR

• Are any of these conditions met?

# Check either one conditional *or* the other
"hello" == "hello" | 2 == 7

[1] TRUE

• Uses | operator
  • Shift + \ on keyboard

Fundamentals: AND

• Are all of the conditions met?

# Are *all* conditions TRUE?
"hello" == "hello" & 2 == 7

[1] FALSE

• Uses & operator
  • Shift + 7 on keyboard

Fundamentals: Summary

EQUAL

• Are two things exactly equal?

"hello" == "hello"

[1] TRUE

• Uses == operator
  • Just two equal signs

OR

• Are any of these conditions met?

"hello" == "hello" | 2 == 7

[1] TRUE

• Uses | operator
  • Shift + \ on keyboard

AND

• Are all of the conditions met?

"hello" == "hello" & 2 == 7

[1] FALSE

• Uses & operator
  • Shift + 7 on keyboard

Practice: Fundamental Conditionals

• We’ll use the base R subset function with the peng_df object
  • If needed, consult the help file for more details (?subset)

• Subset peng_df to only Adelie penguins (and assign to a sub_v1 object)
  • I.e., species == "Adelie"

• How many rows does that subset have?

More Practice: Fundamental Conditionals

• Subset peng_df to Adelie or Gentoo penguins
  • Assign this subset to sub_v2 object

• Subset peng_df to only male Gentoo penguins
  • Assign to sub_v3 object

• How many rows does that subset have?

Discussion: Conditionals

• We’ve covered EQUAL, OR, and AND
  • ==, |, or &

• What unanswered questions do you have?

• What other types of conditional statements would be useful?
  • Think about it in the context of wanting to subset some data

Numeric Conditionals

• For numbers, we can use greater/less than conditionals!

• Greater / less than are expressed as normal
  • > and <

• Adding ‘or equal to’ is done by adding an equal sign
  • >= and <=

Practice: Numeric Conditionals

• Subset peng_df to only penguins with a bill length greater than 40 mm
  • Assign to sub_v7

• Subset peng_df to only penguins with a body mass less than or equal to 4,000 g
  • Assign to sub_v8

Temperature Check

How are you Feeling?

Upcoming Due Dates

Due before lab

(By midnight)

• Muddiest Point #2
• Read Excuse me, do you have a moment to talk about version control?

Due before lecture

(By midnight)

• Homework #2
• Read the British Ecological Society’s Reproducible Code Guide (p.20-25)

Bonus Conditionals

OR with >2 Options

• OR conditionals with many options get cumbersome quickly
  • E.g., x == 1 | x == 2 | x == 3 | x == 4 …

• We can use concatenation and the %in% operator to simplify this!

• %in% is effectively “if any of <this vector> matches the value”
  • E.g., x %in% c(1, 2, 3, 4, …)

Conditionals: NOT

• You can also exclude based on conditions
  • Two different ways of doing this

• For one / a few options: use != for “not equal to”
  • E.g., x != 10

• Can be combined with %in% to exclude a set of options
  • E.g., !x %in% c(...)
  • Note the exclamation mark is before the object

Practice: Advanced Conditionals

• Subset peng_df is species is any of “Adelie”, “Gentoo”, or “Chinstrap”
  • Use the %in% operator

• Subset peng_df to all islands except “Dream”