Knowledge Quiz: Data Frames

Consider the following data frame, df.

   name age height  eyes
1   bob  25    180 green
2  mike  23    178 green
3  mark  24    169  blue
4 steve  26    183 brown

How can we extract the name column and return the selection as a character vector?

df[1]df$namedf[[1]]df[["name"]]df[1:2]

• Wrong: This command will return the name column as a one-column data frame.
• Correct
• Correct
• Correct
• Wrong: This command will return the name and age columns as a \(4 \times 2\) data frame.

Consider the following data frame, df.

   name age height  eyes
1   bob  25    180 green
2  mike  23    178 green
3  mark  24    169  blue
4 steve  26    183 brown

How can we extract the height column and return the selection as a data frame?

df$heightdf[ , "height"]df[2:3]df["height"]df["height", ]

• Wrong: This command will return the height column as a numeric vector.
• Wrong: This command will return the height column as a numeric vector.
• Wrong: This command will return the age and height columns as a \(4 \times 2\) data frame.
• Correct
• Wrong: This command will try to return a row called “height” and fail because there is no such row in df.

Which of the following statements about data frames is correct?

A data frame is a special kind of list, where each slot holds an equal-length vector.A data frame is a special kind of vector, where each element holds and equal-length character vector.A data frame and a matrix are the same underlying object type, just printed differently.A data frame is a special kind of list, where each slot must hold a length-one vector.A data frame has nothing in common with either lists or matrices. It’s a completely separate object type.

A data frame is implemented as a list, where every list slot contains a vector-like object, with the added constraint that all of those vectors must have equal length. This is structure allows us to view data frames as a “rectangular,” table-like structure while still being a list underneath. Data frames are not matrices. Matrices are generalizations of atomic vectors that require every element to share the same type, while data frames are special lists that allow each column to have its own type.

You have a data frame, d, with 10 rows. You run d[1:2] <- 1:3 (attempting to replace the first two columns using a length-3 replacement vector), and R raises an error. Why?

Because you’re trying to replace 20 values, and 3 does not evenly divide 20.Because you’re trying to replace 20 values, and 3 is less than 20.Because you must always use a list or data frame to replace data frame columns.Because 3 is larger than 2 (the number of columns being replaced).Because you can never use a numeric vector to replace data frame columns.

R will recycle a shorter replacement vector to fill a longer selection, but only when the replacement length evenly divides the total number of values needed. Replacing 2 columns across 10 rows requires 20 values, and a length-3 vector doesn’t divide 20 evenly (20 / 3 isn’t a whole number). So, R refuses rather than guessing at an ambiguous fill pattern.

Note

• Click the check-mark button to check your answer.
• Click the question-mark button to see an explanation of the solution.