Packages and Libraries

Why Learn Packages in Data Analytics?

Imagine you are analyzing a massive transaction log. You need to:

1. Filter out cancelled transactions.
2. Group the remaining sales by store region.
3. Compute the average sales amount per region.
4. Draw a beautiful bar chart showing region-wise sales.

If you write this in base R, you have to manage subsetting indices, use complicated aggregate formulas, and write verbose graphing code.

Instead, you want to use a highly readable language for data manipulation like dplyr and a state-of-the-art charting library like ggplot2:

# Leveraging specialized packages (Tidyverse)
library(dplyr)
library(ggplot2)

summary_stats <- transactions %>%
  filter(status == "Completed") %>%
  group_by(region) %>%
  summarize(avg_sales = mean(amount))

ggplot(summary_stats, aes(x = region, y = avg_sales)) + geom_col()

R's core installation is powerful, but its true strength lies in its ecosystem of over 20,000 add-on bundles called Packages. Let's learn how to download, load, and manage packages in R.

1. What is CRAN?

The Comprehensive R Archive Network (CRAN) is a global network of servers that hosts the official R software and thousands of tested R packages.

2. Installing and Loading Packages

Installing: install.packages()

To download a package from CRAN to your computer, use install.packages(). The package name must be inside quotation marks. You only need to run this command once per computer.

# Download and install dplyr from CRAN
install.packages("dplyr")

Loading: library()

Before you can use a package in a script, you must load it into R's memory. The package name does not need to be in quotes. You must run this command in every session/script where you use the package.

# Load the package into memory
library(dplyr)

3. Namespace Resolution: The :: Operator

If you only need to use a single function from a package once, or want to avoid name conflicts (e.g., two packages containing a function named filter), you can call it directly using the double-colon :: operator without loading the package via library():

# Call the filter function directly from the dplyr package namespace
clean_df <- dplyr::filter(customers, Age > 21)

4. Key R Packages for Data Analytics

In R, data analysts heavily rely on a collection of packages called the Tidyverse, designed specifically for data science:

Hands-on Exercises

Exercise 1: Loading and Explicit Namespace Call

Imagine you have a data frame of employee records: employees <- data.frame(Name = c("Alice", "Bob"), Salary = c(50000, 60000)). Write R code to:

1. Load the dplyr package using library().
2. Extract the Salary column as a vector using dplyr::pull(employees, Salary).
3. Print the resulting salary vector.

# Write your code below and click Run Code

# Load package
library(dplyr)

employees <- data.frame(Name = c("Alice", "Bob"), Salary = c(50000, 60000))

# Call function with namespace
salaries <- dplyr::pull(employees, Salary)
print(salaries) # Output: 50000 60000

Exercise 2: Tidyverse Inspection

R has a meta-package called tidyverse that loads all key data science packages at once. Write R code to:

1. Load the tidyverse library. (In our interactive editor, this is pre-installed).
2. Use the package function stringr::str_to_title("r programming language") to convert the string.
3. Print the title-cased result.

# Write your code below and click Run Code

library(tidyverse)

title_string <- stringr::str_to_title("r programming language")
print(title_string) # Output: "R Programming Language"