Assignment 3
Kento Fujikawa | Student ID: 101315119
2026-07-17
1
library(dplyr)
library(tidyr)
library(ggplot2)
data(mpg)1 a)
mpg_clean <- mpg %>%
select(manufacturer, model, year, cyl, displ, cty, hwy, class, drv) %>%
mutate(
avg_mpg = round((cty + hwy) / 2, 1),
efficiency_rating = case_when(
avg_mpg < 20 ~ "Low",
avg_mpg >= 20 & avg_mpg <= 25 ~ "Medium",
avg_mpg > 25 ~ "High"
),
engine_size = case_when(
displ < 2.5 ~ "Small",
displ >= 2.5 ~ "Large"
)
)
head(mpg_clean, 5)## # A tibble: 5 × 12
## manufacturer model year cyl displ cty hwy class drv avg_mpg
## <chr> <chr> <int> <int> <dbl> <int> <int> <chr> <chr> <dbl>
## 1 audi a4 1999 4 1.8 18 29 compact f 23.5
## 2 audi a4 1999 4 1.8 21 29 compact f 25
## 3 audi a4 2008 4 2 20 31 compact f 25.5
## 4 audi a4 2008 4 2 21 30 compact f 25.5
## 5 audi a4 1999 6 2.8 16 26 compact f 21
## # ℹ 2 more variables: efficiency_rating <chr>, engine_size <chr>dim(mpg_clean)## [1] 234 121 b)
subset A
subset_A <- mpg_clean %>%
filter(efficiency_rating == "High",
engine_size == "Small")
subset_A## # A tibble: 27 × 12
## manufacturer model year cyl displ cty hwy class drv avg_mpg
## <chr> <chr> <int> <int> <dbl> <int> <int> <chr> <chr> <dbl>
## 1 audi a4 2008 4 2 20 31 compact f 25.5
## 2 audi a4 2008 4 2 21 30 compact f 25.5
## 3 chevrolet malibu 2008 4 2.4 22 30 midsize f 26
## 4 honda civic 1999 4 1.6 28 33 subcompact f 30.5
## 5 honda civic 1999 4 1.6 24 32 subcompact f 28
## 6 honda civic 1999 4 1.6 25 32 subcompact f 28.5
## 7 honda civic 1999 4 1.6 23 29 subcompact f 26
## 8 honda civic 1999 4 1.6 24 32 subcompact f 28
## 9 honda civic 2008 4 1.8 26 34 subcompact f 30
## 10 honda civic 2008 4 1.8 25 36 subcompact f 30.5
## # ℹ 17 more rows
## # ℹ 2 more variables: efficiency_rating <chr>, engine_size <chr>nrow(subset_A)## [1] 27subset B
subset_B <- mpg_clean %>%
filter(class %in% c("compact",
"subcompact",
"midsize"))
subset_B## # A tibble: 123 × 12
## manufacturer model year cyl displ cty hwy class drv avg_mpg
## <chr> <chr> <int> <int> <dbl> <int> <int> <chr> <chr> <dbl>
## 1 audi a4 1999 4 1.8 18 29 compact f 23.5
## 2 audi a4 1999 4 1.8 21 29 compact f 25
## 3 audi a4 2008 4 2 20 31 compact f 25.5
## 4 audi a4 2008 4 2 21 30 compact f 25.5
## 5 audi a4 1999 6 2.8 16 26 compact f 21
## 6 audi a4 1999 6 2.8 18 26 compact f 22
## 7 audi a4 2008 6 3.1 18 27 compact f 22.5
## 8 audi a4 quattro 1999 4 1.8 18 26 compact 4 22
## 9 audi a4 quattro 1999 4 1.8 16 25 compact 4 20.5
## 10 audi a4 quattro 2008 4 2 20 28 compact 4 24
## # ℹ 113 more rows
## # ℹ 2 more variables: efficiency_rating <chr>, engine_size <chr>nrow(subset_B)## [1] 123subset C
subset_C <- mpg_clean %>%
filter(manufacturer %in% c("honda", "toyota"),
year == 2008)
subset_C## # A tibble: 18 × 12
## manufacturer model year cyl displ cty hwy class drv avg_mpg
## <chr> <chr> <int> <int> <dbl> <int> <int> <chr> <chr> <dbl>
## 1 honda civic 2008 4 1.8 26 34 subc… f 30
## 2 honda civic 2008 4 1.8 25 36 subc… f 30.5
## 3 honda civic 2008 4 1.8 24 36 subc… f 30
## 4 honda civic 2008 4 2 21 29 subc… f 25
## 5 toyota 4runner 4wd 2008 6 4 16 20 suv 4 18
## 6 toyota 4runner 4wd 2008 8 4.7 14 17 suv 4 15.5
## 7 toyota camry 2008 4 2.4 21 31 mids… f 26
## 8 toyota camry 2008 4 2.4 21 31 mids… f 26
## 9 toyota camry 2008 6 3.5 19 28 mids… f 23.5
## 10 toyota camry solara 2008 4 2.4 21 31 comp… f 26
## 11 toyota camry solara 2008 4 2.4 22 31 comp… f 26.5
## 12 toyota camry solara 2008 6 3.3 18 27 comp… f 22.5
## 13 toyota corolla 2008 4 1.8 28 37 comp… f 32.5
## 14 toyota corolla 2008 4 1.8 26 35 comp… f 30.5
## 15 toyota land cruiser … 2008 8 5.7 13 18 suv 4 15.5
## 16 toyota toyota tacoma… 2008 4 2.7 17 22 pick… 4 19.5
## 17 toyota toyota tacoma… 2008 6 4 15 18 pick… 4 16.5
## 18 toyota toyota tacoma… 2008 6 4 16 20 pick… 4 18
## # ℹ 2 more variables: efficiency_rating <chr>, engine_size <chr>nrow(subset_C)## [1] 18subset D
subset_D <- mpg_clean %>%
filter(cyl >= 6,
avg_mpg < 20)
subset_D## # A tibble: 101 × 12
## manufacturer model year cyl displ cty hwy class drv avg_mpg
## <chr> <chr> <int> <int> <dbl> <int> <int> <chr> <chr> <dbl>
## 1 audi a6 quattro 1999 6 2.8 15 24 mids… 4 19.5
## 2 audi a6 quattro 2008 8 4.2 16 23 mids… 4 19.5
## 3 chevrolet c1500 suburba… 2008 8 5.3 14 20 suv r 17
## 4 chevrolet c1500 suburba… 2008 8 5.3 11 15 suv r 13
## 5 chevrolet c1500 suburba… 2008 8 5.3 14 20 suv r 17
## 6 chevrolet c1500 suburba… 1999 8 5.7 13 17 suv r 15
## 7 chevrolet c1500 suburba… 2008 8 6 12 17 suv r 14.5
## 8 chevrolet corvette 1999 8 5.7 15 23 2sea… r 19
## 9 chevrolet corvette 2008 8 7 15 24 2sea… r 19.5
## 10 chevrolet k1500 tahoe 4… 2008 8 5.3 14 19 suv 4 16.5
## # ℹ 91 more rows
## # ℹ 2 more variables: efficiency_rating <chr>, engine_size <chr>nrow(subset_D)## [1] 101Subset A mainly contains smaller engines with high fuel efficiency, while Subset D contains larger engines with at least six cylinders and poor fuel economy. Overall, larger engines and higher cylinder counts tends to reduce fuel efficiency, whereas smaller engines generally achieve better fuel economy.
1 c)
i)
manufacturer_summary <- mpg_clean %>%
group_by(manufacturer) %>%
summarise(
mean_mpg = round(mean(avg_mpg), 2),
mean_displ = round(mean(displ), 2),
num_models = n()
) %>%
arrange(desc(mean_mpg)) %>%
slice_head(n = 5)
manufacturer_summary## # A tibble: 5 × 4
## manufacturer mean_mpg mean_displ num_models
## <chr> <dbl> <dbl> <int>
## 1 honda 28.5 1.71 9
## 2 volkswagen 25.1 2.26 27
## 3 hyundai 22.8 2.43 14
## 4 subaru 22.4 2.46 14
## 5 audi 22.0 2.54 18ii)
class_year_summary <- mpg_clean %>%
group_by(class, year) %>%
summarise(
median_hwy = median(hwy),
median_cty = median(cty),
.groups = "drop"
) %>%
arrange(class, year)
class_year_summary## # A tibble: 14 × 4
## class year median_hwy median_cty
## <chr> <int> <dbl> <dbl>
## 1 2seater 1999 24.5 15.5
## 2 2seater 2008 25 15
## 3 compact 1999 26 19
## 4 compact 2008 29 20.5
## 5 midsize 1999 26 18
## 6 midsize 2008 28 19
## 7 minivan 1999 22 16
## 8 minivan 2008 23 16
## 9 pickup 1999 17 13
## 10 pickup 2008 17 13
## 11 subcompact 1999 26 19
## 12 subcompact 2008 26.5 18.5
## 13 suv 1999 17 14
## 14 suv 2008 18 13iii)
drive_summary <- mpg_clean %>%
group_by(drv) %>%
summarise(
mean_hwy = round(mean(hwy), 2),
sd_hwy = round(sd(hwy), 2),
min_cty = min(cty),
max_cty = max(cty)
)
drive_summary## # A tibble: 3 × 5
## drv mean_hwy sd_hwy min_cty max_cty
## <chr> <dbl> <dbl> <int> <int>
## 1 4 19.2 4.08 9 21
## 2 f 28.2 4.21 11 35
## 3 r 21 3.66 11 18Front-wheel drive (f) has the highest average highway mileage (28.16 mpg) while also having the highest standard deviation (4.21 mpg), meaning highway fuel economy varies the most among front-wheel drive vehicles. For buyers, this indicates that front-wheel drive is generally the most fuel-efficient option for highway driving, but fuel economy can differ considerably between individual models.
2
2 a)
ggplot(mpg,
aes(x = displ,
y = hwy,
colour = class,
size = cyl)) +
geom_point(shape = 16,
alpha = 0.6) +
labs(
title = "Engine Displacement vs Highway Mileage",
x = "Engine Displacement (Litres)",
y = "Highway Mileage (MPG)",
colour = "Vehicle Class",
size = "Cylinders"
) +
theme_minimal()
The scatter plot shows a negative relationship between engine displacement and highway mileage. As engine displacement increases, highway fuel economy generally decreases. SUVs and pickup trucks tend to have the largest engines and the lowest highway mileage because they are heavier vehicles that require more fuel.
2 b)
ggplot(mpg,
aes(x = drv,
y = cty,
fill = drv)) +
geom_boxplot() +
facet_wrap(~year) +
scale_fill_manual(values = c(
"4" = "steelblue",
"f" = "forestgreen",
"r" = "tomato"
)) +
labs(
title = "City Mileage by Drive Train and Year",
x = "Drive Train",
y = "City Mileage (MPG)",
fill = "Drive Train"
) +
theme_bw()
The median city mileage for each drivetrain is similar between 1999 and 2008, although front-wheel drive vehicles generally achieve the highest city mileage. Overall, fuel economy appears to remain stable across the two years, with only small improvements for some drivetrain types.
2 c)
mpg %>%
filter(year == 2008) %>%
count(manufacturer) %>%
ggplot(aes(
x = reorder(manufacturer, n),
y = n,
fill = manufacturer
)) +
geom_col() +
geom_text(aes(label = n),
hjust = -0.1) +
coord_flip() +
labs(
title = "Number of 2008 Vehicles by Manufacturer",
x = "Manufacturer",
y = "Number of Vehicles"
) +
theme_bw() +
guides(fill = "none")
The manufacturer with the longest horizontal bar has the greatest number of vehicle records in the 2008 dataset. Lincoln is the only one with a single record
2 d)
ggplot(mpg,
aes(x = cty,
y = hwy)) +
geom_point(
colour = "steelblue",
alpha = 0.6
) +
geom_smooth(
method = "lm",
colour = "red",
fill = "pink"
) +
geom_abline(
slope = 1,
intercept = 0,
linetype = "dashed",
colour = "grey50"
) +
annotate(
"text",
x = 13,
y = 13,
label = "hwy = cty",
colour = "black"
) +
labs(
title = "City Mileage vs Highway Mileage",
x = "City Mileage (MPG)",
y = "Highway Mileage (MPG)"
) +
theme_classic()
Most data points lie above the dashed reference line, meaning that
highway mileage is generally higher than city mileage for the same
vehicle. Since vehicles consume less fuel during steady highway driving
compared o stop and go city driving, this outcome was expected. The
regression line fits the data well because the points follow a strong
positive linear trend with relatively little scatter around the fitted
line.