STAT Homework 3 Solution

Let’s work on the mpg dataset in the package ggplot2. You can use the following code to load the data. Use necessary code to read the description of the dataset, which contains 234 samples and 11 variables.


    • Install the package if you never did install . packages (" ggplot 2" )

    • Load  the  pacakge

library ( ggplot 2 )

    • Load the mpg dataset data ("mpg")



Problem 1 (1    6 = 6 points)

Let’s first clean the data:

    (a) Turn the variable cyl to an ordered factor variable with levels “4”, “5”, “6”, and “8”

    (b) Turn the variable trans to a factor variable, of which unique values are “auto” and “manu” (Hint: use the function substr() to extract substrings in a character vector before converting to a factor vector)

    (c) Turn the variable drv to an ordered factor variable with levels “f”, “r”, and “4”,

    (d) Turn the variable fl to a factor variable, of which unique values are “gasoline”, “diesel”, and “other” (Hint, “other” should include “e” and “c” in the original vari-able: “e” for E85, which is an ethanol fuel blend of 85% ethanol fuel and 15% gasoline and “c” for compressed natural gas)

    (e) Turn the variable class to an ordered factor variable with levels “2seater”, “subcom-pact”, “compact”, “midsize”, “suv”, “minivan”, and “pickup”



1

Country
Manufacturer


United States
Chevrolet, Dodge, Ford, Jeep, Lincoln, Mercury, Pontiac


Japan
Honda, Nissan, Subaru, Toyota


Germany
Audi, Volkswagen


South Korea
Hyundai


Great Britain
Land Rover




    (f) Create a new variable of country to indicate the manufacturer base location (Hint: You can refer to the above table)

Hint: You should get the following response after applying the function str() on the cleaned dataset

$
m a n u f a c t u r e r :  chr
" audi
"
" audi "
" audi "  " audi "  ...



$
model



:  chr
"a4
"

"a4"  "a4"
"a4"  ...





$
displ



:  num
1 .8

1
.8222.
8  2.8  3.1
1.8  1.8  2  ...

$
year




:  int
1999

1999
2008
2008  1999
1999
2008
199

9
1999
2008  ...















$
cyl




:  Factor
w /

4
levels
"4","5","6","8":
1
1
1
1

3
3

3
1
1
1  ...














$
trans



:  Factor
w /

2
levels
" auto " ," manu ":  1
2
2
1
1

2
1

2
1
2  ...















$
drv




:  Factor
w /

3
levels
"f" ,"r" ,"4":  1  1
1
1
1
1

1
3

3
3  ...
















$
cty




:  int
18

21

20
21
16
18  18  18
16  20  ...



$
hwy




:  int
29

29

31
30
26
26  27  26
25  28  ...



$
fl





:  Factor
w /

3
levels
" diesel " ," gasoline " ,..:


2
2
2
2
2
2
2
2
2
2  ...










$
class



:
Factor
w /

7
levels
" 2 seater " ," s u b c o m p a c t "


,..:

3
3
3
3
3
3
3  3
3
3  ...







$
country


:  chr
" germany "
" germany "  " germany "
" germany

"  ...














































Problem 2 (1    4 = 4 points)

Let’s analyze the data. For each figure, please make it complete/readable, in other words, it should include all the label information, title, and legend if necessary.

    (a) Draw a bar plot of the variable country and arrange the country in decreasing order in terms of the number of samples. Which country has the most samples in this dataset? Which has the least?


2
    (b) Summarize what a typical U.S. car looks like, in terms of engine displacement (i.e. displ), number of cylinders (i.e. cyl), type of transmission (i.e. trans), drive type (i.e. drv), fuel type (i.e. fl), and type of car (i.e. class)? (Hint: Use the function table() to find the mode for each of the above discrete univariate data)

    (c) Make a boxplot of the combined miles per gallon (i.e. (cty + hwy)/2) of U.S. cars and Japan cars, respectively, and report their means, medians, standard deviations, and IQRs.

    (d) Make a histogram of the engine displacement (i.e. displ) of U.S. cars and Japan cars, respectively, and describe their shapes.

















































3