Exercises Part 2: Data Frames and Other Data Structures

Note: new exercises coming soon.

Exercise: Explore Differences in Selecting from data.frame

Make a data frame, called mydata, that has 3 variables: var1 which has the numbers 1 to 10, var2 which has the letters A-J, and var3 which has any 10 numbers you want to include.

After you’ve made the data frame, then add a fourth column, month, with the first 10 months:

mydata<-data.frame(var1=1:10, var2=LETTERS[1:10], 
                   var3=c(9,4.5,33,45.6,-14,3,7,0.2,0,7))
mydata$month <- month.name[1:10]
mydata

##    var1 var2  var3     month
## 1     1    A   9.0   January
## 2     2    B   4.5  February
## 3     3    C  33.0     March
## 4     4    D  45.6     April
## 5     5    E -14.0       May
## 6     6    F   3.0      June
## 7     7    G   7.0      July
## 8     8    H   0.2    August
## 9     9    I   0.0 September
## 10   10    J   7.0   October

Run each command below to figure out what type of data you get back.

Hint: Use the function typeof() to examine what is returned in each case.

mydata[1]
mydata[[1]]
mydata$var2
mydata["var2"]
mydata[1, 1]
mydata[, 1]
mydata[1,]
mydata[-1,]

Exercise: Working with a data.frame

R has some built-n data sets. One of them is called iris. You can just use the iris object (it’s a data.frame) without creating it first.

Hint: if you want iris to show up in the Environment tab, load it into your environment with data(iris). Otherwise, you can still use it, but it may not show in that tab.

Get the dimensions of iris. Then get a list of the names.

Output the first 10 rows of iris.

View the iris data frame in the RStudio data viewer.

Select from the iris data frame the observations where the Sepal.Width is less than 2.5. Do the same, but add the condition that the Sepal.Length must also be less than 5.

Using which.max, select a row with maximum Sepal.Length. Is there only one row with the maximum value?

Rename the column Petal.Width to petalwidth. Challenge: do this without hard coding in the column number.

Exercise: Working with a data.frame: mtcars

Another built-in data set is called mtcars.

Explore mtcars. Look at the help page for mtcars to see the variable definitions.

• Get the cars with 6 cylinders
• Get just the weight (wt) column
• Make a new data frame called fuel_efficient cars that has cars with mpg > 25
• Do the fuel efficient cars have lower average horsepower (hp) than the overall average?
• Challenge: Which car with 3 gears has the highest horsepower (hp)? Hint: subset first, save as new variable, then find max

Challenge Exercise: Use subset

Look up the help for the subset function. Can you use it to repeat some of the operations in the two exercises above?

Note: we’ll cover this function later in the workshop.

Exercise: Select elements from a list

The repurrrsive package has example list objects in it. Load the package (install first if needed). Then:

• View the sw_people list with the View function. What’s this a list of?
• Get the contents of first element in the list
• Get the name of the first person in the list
• What’s the gender of the last person in the list?

Exercise: Fix subsetting errors

A researcher is trying to select observations from the iris data frame for the setosa species, but gets an error:

iris[Species="setosa"]

## Error in `[.data.frame`(iris, Species = "setosa"): unused argument (Species = "setosa")

Correct this expression to select the observations the researcher wants. Hint: there may be more than 1 thing wrong.

Exercise: Counting with Conditionals

Using the built-in data set mtcars:

• How many cars have mpg greater than 30?
  
• How many have horse power (hp) less than 100?

Hint: you can use the sum function to count the number of TRUE observations in a vector (as we did with sum(is.na(x))). This works because as.numeric(TRUE) == 1 and as.numeric(FALSE) == 0.

Challenge Exercise: Matrix Manipulation

Using information from Linear Algebra in R by Søren Højsgaard if needed, create a 10 x 5 matrix of random normal numbers and a 5 x 1 vector of values 1 to 5 and multiply them using matrix multiplication. Transpose your result.

Create a 6 x 6 matrix of random normal draws and take it’s inverse (solve it). Extract the diagonal from the result.

Hint: Generate random normal draws with rnorm: look it up to see the options.

Hint: Matrix multiplication operator is %*%.

Challenge Exercise: List Indexing

Select the vector of letters from the following object.

Do it both using names (hint: printing the object might give you a hint) and using index numbers.

Extra hint: use the object viewer in RStudio to view the list after you create it. See if there’s any functionality to help you with this challenge.

nested_list <- list(level1 = list(level2 = list(letters = LETTERS)))

Challenge Exercise: Use apply

Use the apply function to get the average value of each variable in mtcars. You’ll need to look at the help page to figure out how to use it.