Solutions: Basic Data Manipulation

Materials adapted from Adrien Osakwe, Larisa M. Soto and Xiaoqi Xie. 
library(gapminder)

Writing data

Write a data processing snippet to include only the data points collected after 1995 in Asian countries as a CSV file

asia<-gapminder[gapminder$year > 1995 & gapminder$continent=="Asia", ]
write.table(asia,
            file = "../data/gapminder_after1995_asia.csv",
            sep = ",", 
            quote = FALSE, 
            row.names = FALSE)

Separate the gapminder data frame into 5 individual data frames, one for each continent. Store those 5 data frames as an RData file in the objects folder called continents.RData.

asia<-gapminder[gapminder$continent=="Asia", ]
africa<-gapminder[gapminder$continent=="Africa", ]
oceania<-gapminder[gapminder$continent=="Oceania", ]
europe<-gapminder[gapminder$continent=="Europe", ]
americas<-gapminder[gapminder$continent=="Americas", ]

save(asia,africa,oceania,europe,americas,file="../data/continents.RData")
Note

Understanding File Paths: The Tree Analogy

When you are working in an RStudio Project, your Working Directory is your “Home Base.” Think of your folders as a tree:

  • ./ (The Current Folder): This is where your .qmd or .Rmd file is currently saved.

  • ../ (The Parent Folder): This tells R to move up one level in the folder tree.

  • ../../ (The Grandparent Folder): This moves up two levels.

Breakdown of "../data/continents.RData"

When you run that code, you are telling R to follow these exact steps:

  1. ../: Leave the current folder (e.g., the scripts or vignettes folder).

  2. data/: Look for a folder named data at that higher level.

  3. continents.RData: Save the file inside that folder with this specific name.

Note

The “Folder Not Found” Error

R is a great calculator, but it isn’t a folder manager. If you tell R to save a file in a folder named data/ but that folder doesn’t exist yet, R will throw an error: No such file or directory.

Before saving, you must create the folder manually or use this code:

# Create a 'data' folder in your current working directory
dir.create("data")

Exploring data frames

Finish exploring the gapminder data frame and:

  • Find the number of rows and the number of columns
  • Print the data type of each column
  • Explain the meaning of everything that str(gapminder) prints
dim(gapminder)

typeof(gapminder$country)
typeof(gapminder$continent)
typeof(gapminder$year)
typeof(gapminder$lifeExp)
typeof(gapminder$pop)
typeof(gapminder$gdpPercap)

str(gapminder)
Note

Column Data Types

While typeof() gives you the low-level R storage type, in data analysis, we often care more about the class (how R treats the data).

Column typeof() result Interpretation
country integer Stored as numbers (1, 2, 3) because it is a Factor.
continent integer Also a Factor (categorical data).
year integer Discrete whole numbers.
lifeExp double Floating-point numbers (decimals).
pop integer Whole numbers (count data).
gdpPercap double Floating-point numbers (decimals). 
class(gapminder$country)
class(gapminder$continent)
class(gapminder$year)
class(gapminder$lifeExp)
class(gapminder$pop)
class(gapminder$gdpPercap)
Note

The str() Output

The str() (Structure) function is arguably the most useful command in Base R. It provides a compact summary of any object.

The Header

tibble [1,704 × 6] (S3: tbl_df/tbl/data.frame)

  • tibble / data.frame: This tells you the object “flavor.” A tibble is a modern, user-friendly version of a data frame used in the Tidyverse.

  • [1,704 x 6]: Confirms the dimensions (Rows x Columns).

The Column Details

  • $ country : Factor w/ 142 levels:

    • Factor: This is Categorical data.

    • 142 levels: There are 142 unique countries.

    • 1 1 1...: Internally, R stores “Afghanistan” as the number 1 to save memory.

  • $ year : int [1:1704]:

    • int: Integer. Whole numbers only.

    • [1:1704]: This column is a vector with indices ranging from 1 to 1,704.

  • $ lifeExp : num:

    • num: Numeric. This is a “double” or decimal number. This is what you use for most statistical calculations.

In which years has the GDP of Canada been larger than the average of all data points?

canada<-gapminder[gapminder$country=="Canada",]
mgdp<-mean(canada$gdpPercap)
canada[canada$gdpPercap>mgdp,"year"]

Find the mean life expectancy of Switzerland before and after 2000

swiss<-gapminder[gapminder$country=="Switzerland",]
mean(swiss[swiss$year<2000,]$lifeExp) # Before
mean(swiss[swiss$year>2000,]$lifeExp) # After

You discovered that all the entries from 2007 are actually from 2008. Create a copy of the full gapminder data frame in an object called gp. Then change the year column to correct the entries from 2007.

gp<-gapminder
gp[gp$year==2007,"year"]<-2008
gp[gp$year==2008,]

Bonus - Find the mean life expectancy and mean gdp per continent using the function tapply

tapply(gapminder$lifeExp,gapminder$continent,mean)