In kate-pyper/MM916ProgrammingExercises:

library(learnr)
library(testwhat)
knitr::opts_chunk$set(echo = FALSE)
tutorial_options(exercise.timelimit = 60, exercise.checker=testwhat::testwhat_learnr)

require(tidyverse)
data(mtcars)

For the following examples you should use functions from dplyr to extract the relevant pieces of information from the mtcars data set:

head(mtcars)

Exercise 1

Find the average miles per gallon (mpg) for cars after distinguishing between the number of gears (gear) that a car has.

# tidy the penguins data
mtcars %>% group_by(gear) %>% summarise(Average=mean(mpg))

ex() %>% {
  check_error(.)
  check_function(., "group_by", not_called_msg="You should use group_by to find a grouped summary") %>% check_result(.) %>% check_equal(.)
  check_function(., "summarise") %>% check_result(.) %>% check_equal(., incorrect_msg="Check your grouping variable, and make sure that you have used the mean function correctly")
}

question("Which value of gears had the lowest mpg?",
  answer("3", correct=TRUE),
  answer("4"),
  answer("5"), 
  correct="Yes - Cars with fewer gears are less fuel efficient!"
)

Exercise 2

Find the median horse-power (hp) after taking into account engine type (vs) and transmission type (am)

mtcars %>% group_by(vs, am) %>% summarise(Med_hp=median(hp))

ex() %>%{
  check_error(.)
  check_or(.,
           {check_function(., "group_by") %>% check_result(.) %>% check_equal(., incorrect_msg="Check the way you have grouped the data. Does this match what is in the question?")
             check_function(., "median")
  check_function(., "summarise") %>% check_result(.) %>% check_equal(.)},
  {override_solution_code(., "mtcars %>% group_by(am, vs) %>% summarise(Med_hp=median(hp))") %>% {check_function(., "group_by") %>% check_result(.) %>% check_equal(., incorrect_msg="Check the way you have grouped the data. Does this match what is in the question? Have you calculated the correct summary?")
  check_function(., "summarise") %>% check_result(.) %>% check_equal(.)}})
}

Exercise 3

Use the quantile() function to extract the 25th and 75th percentiles (lower and upper quartiles) of the weight column (wt) after grouping by the number of cylinders (cyl).

mtcars %>% group_by(cyl) %>% summarise(Quartiles=quantile(wt, c(0.25, 0.75)))

ex() %>% {
  check_error(.)
  check_function(., "group_by", not_called_msg="You should use group_by to find a grouped summary") %>% check_result(.) %>% check_equal(.)
  check_function(., "summarise") %>% check_result(.) %>% check_equal(., incorrect_msg="Make sure that you have grouped by the correct variable. Have you specified the correct probabilities in the quantile function?")
}

question("It can be a bad idea to do this sort of summary...why?",
  answer("The quantile function returns multiple values and so is not compatible with the summary function"),
  answer("The output is not labelled so it is easy to mix up the values", correct=TRUE),
  answer("It isn't compatible with tidy data principles"), 
  correct="Yes - in this case it is clear that the lower of each pair is the lower quartile, but it is not always so clear cut"
)

Exercise 4

Find the variance of the columns mpg, disp, hp, drat, wt and qsec (i.e. the continuous variables.)

mtcars %>% summarise(across(c(mpg, disp:qsec), var))

ex() %>% {
  check_error(.)
  check_function(., "across", not_called_msg="The easiest way to do these computations is to use the across function")
  check_function(., "summarise") %>% check_result(.) %>% check_equal(., incorrect_msg="Have you selected the correct variables?")
}