Session 6: Factors

Factors form the basis for many powerful operations in R, including many performed on tabular data. The motivation for factors comes from the notion of categorical variables. These variables are non-numeric in nature corresponding to categories such as male and female, or Democrat, Republican and Independent.

A factor might be viewed simply as a vector with a bit of more information added. The extra information consists of a record of distinct values in that vector, which are called: levels.

Let us look at some examples of factors. We will make use of the package forcats, which is one of the 8 core tidyverse packages. Therefore, we start by loading the tidyverse:

library(tidyverse)

#> ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.0 ──

#> ✔ ggplot2 3.3.3     ✔ purrr   0.3.4
#> ✔ tibble  3.0.4     ✔ dplyr   1.0.2
#> ✔ tidyr   1.1.2     ✔ stringr 1.4.0
#> ✔ readr   1.4.0     ✔ forcats 0.5.0

#> ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
#> ✖ dplyr::filter() masks stats::filter()
#> ✖ dplyr::lag()    masks stats::lag()


## Check whether "forcats" is listed among the loaded packages.
## Alternatively, you could load "forcats" (and "ggplot2") separately:
# install.packages("forcats")
# library(forcats)
# library(ggplot2)

Example 1: From a numeric vector to a factor

Let us create a factor xf from a vector x with the numbers 5, 12, 13, and 12:

x <- c(5,12, 13,12)
x

#> [1]  5 12 13 12


# Convert the vector to a factor:
xf <- factor(x)
xf

#> [1] 5  12 13 12
#> Levels: 5 12 13

The distinct values in xf are 5, 12 and 13, and are listed as levels.

Let us look in a bit more details at our factor using the R functions str and unclass:

str(xf)

#>  Factor w/ 3 levels "5","12","13": 1 2 3 2

unclass(xf)

#> [1] 1 2 3 2
#> attr(,"levels")
#> [1] "5"  "12" "13"

Notice that the values in the factor are not stored as (5, 12, 13, 12), but rather as (1, 2, 3, 2)!

This means that our data consists first of a level-1 value, then level-2 and level 3 values, and finally another level-2 value. So, the data has been recorded by level.

The values attached to each level are recorded too, but as characters such as "5" rather than as numbers such as 5.

Example 2: From a character vector to a factor

We will use the levels Democrat (D), Republican (R), and Independent (I). First, we save a vector:

y <- c("D", "R", "R", "I", "R", "I", "D", "I")
y

#> [1] "D" "R" "R" "I" "R" "I" "D" "I"

str(y)

#>  chr [1:8] "D" "R" "R" "I" "R" "I" "D" "I"

Then, we again convert the vector to a factor, and look at the levels:

fy <- factor(y)
fy

#> [1] D R R I R I D I
#> Levels: D I R

unclass(fy)

#> [1] 1 3 3 2 3 2 1 2
#> attr(,"levels")
#> [1] "D" "I" "R"

Example 3: Ordering factor levels

Some variables can be challenging to sort automatically, because the desired sorting order is not alphabetical or numeric.

For instance, months that are listed using characters:

months_vector <- c("Dec", "Apr", "Jan", "Mar")

# Try to sort using the `sort` function
sort(months_vector)

#> [1] "Apr" "Dec" "Jan" "Mar"

That didn’t sort in a useful way. But, the problem can be fixed by using a factor.

First, we create a list of the valid levels, which are all 12 months in a year:

month_levels <- c("Jan", "Feb", "Mar", "Apr", "May", "Jun",
                  "Jul", "Aug", "Sep", "Oct", "Nov", "Dec")

Then we convert the vector into a factor, like before, but now we additionally specify the desired levels of the factor, in order, using the levels argument:

months_factor <- factor(months_vector, levels = month_levels)

Now it sorts the way we want to!

sort(months_factor)

#> [1] Jan Mar Apr Dec
#> Levels: Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Example 4: Use of factors in plots with forcats

4A: Plot after reordering manually with fct_relevel()

We will use the mtcars data, which was extracted from the 1974 Motor Trend US magazine, and comprises fuel consumption and 10 aspects of automobile design and performance for 32 automobiles (1973–74 models) – a data frame with 32 observations for 11 (numeric) variables,

data(mtcars)

names(mtcars)

#>  [1] "mpg"  "cyl"  "disp" "hp"   "drat" "wt"   "qsec" "vs"   "am"   "gear"
#> [11] "carb"

dim(mtcars)

#> [1] 32 11

str(mtcars)

#> 'data.frame':  32 obs. of  11 variables:
#>  $ mpg : num  21 21 22.8 21.4 18.7 18.1 14.3 24.4 22.8 19.2 ...
#>  $ cyl : num  6 6 4 6 8 6 8 4 4 6 ...
#>  $ disp: num  160 160 108 258 360 ...
#>  $ hp  : num  110 110 93 110 175 105 245 62 95 123 ...
#>  $ drat: num  3.9 3.9 3.85 3.08 3.15 2.76 3.21 3.69 3.92 3.92 ...
#>  $ wt  : num  2.62 2.88 2.32 3.21 3.44 ...
#>  $ qsec: num  16.5 17 18.6 19.4 17 ...
#>  $ vs  : num  0 0 1 1 0 1 0 1 1 1 ...
#>  $ am  : num  1 1 1 0 0 0 0 0 0 0 ...
#>  $ gear: num  4 4 4 3 3 3 3 4 4 4 ...
#>  $ carb: num  4 4 1 1 2 1 4 2 2 4 ...

we will select six variables (mpg, cyl, disp, hp, and wt) to create a dataset Data.

• mpg: Miles per (US) gallon,
• cyl: Number of cylinders
• disp: Displacement (cu.in.)
• hp: Horse power
• wt: Weight (in 1000 lbs)

Data <- mtcars %>%
  select("mpg", "cyl", "disp", "hp", "wt")

Now, we’ll add a new column cyl_chr by converting cyl from numeric to character:

Data <- Data %>%
  mutate(cyl_chr = recode(cyl,`4` = "Four", `6` = "Six", `8` = "Eight"))

head(Data)

#>    mpg cyl disp  hp    wt cyl_chr
#> 1 21.0   6  160 110 2.620     Six
#> 2 21.0   6  160 110 2.875     Six
#> 3 22.8   4  108  93 2.320    Four
#> 4 21.4   6  258 110 3.215     Six
#> 5 18.7   8  360 175 3.440   Eight
#> 6 18.1   6  225 105 3.460     Six

We plot a bar chart for cyl_chr:

Data %>%
  ggplot(aes(x = cyl_chr)) +
  geom_bar() 

In the plot, the levels of the factor were arranged in alphabetical order (Eight, Four, and Six).

Instead, we want the bar graph arranged in the order Four, Six, and Eight.

An alternative to using factor(levels = ...) like we did above, is to use the fct_relevel() function from the forcats package:

Data %>%
  mutate(cyl_chr = fct_relevel(cyl_chr, "Four", "Six", "Eight")) %>%
  ggplot(aes(x = cyl_chr)) +
  geom_bar() +
  labs(x = "Cylinder", y = "Number of cars")

4B: Plot after reordering by the value of another column (fct_reorder)

Create a dataset called Data_a:

Data_a <- data.frame(name = c("North", "South", "East", "West"),
                     var = sample(seq(1, 10), 4))

Plot a bar chart of Data_a:

Data_a %>%
  ggplot(aes(x = name, y = var)) +
  geom_bar(stat = "identity", fill = "#f68034", alpha = 0.6, width = 0.4) 

Reorder following the value of another column using the fct_reorder() function, and flip the plot:

Data_a %>%
  mutate(name = fct_reorder(name, var)) %>%
  ggplot(aes(x = name, y = var)) +
  geom_bar(stat = "identity", fill = "#f68034", alpha = 0.6, width = 0.4) +
  coord_flip() 

There are several more convenient reordering functions in the forcats package, including:

• fact_infreq() to reorder by occurrence frequencies of each level (see the picture at the top of the post).

• fct_inorder() to reorder by order of appearance in the dataframe. This can be useful, for example, if your dataframe has already been sorted properly, and you just need to prevent automatic alphabetic reordering when plotting.

Breakout rooms!

For the Breakout room exercises, we will use datasets from mtcars and the gss_cat dataset from the forcats package.

Exercise 1

data("mtcars")

gear_df <- mtcars %>%
  mutate(gear_chr = recode(gear,
                         `3`= "Three",
                         `4` ="Four",
                         `5`= "Five"))

head(gear_df)

#>    mpg cyl disp  hp drat    wt  qsec vs am gear carb gear_chr
#> 1 21.0   6  160 110 3.90 2.620 16.46  0  1    4    4     Four
#> 2 21.0   6  160 110 3.90 2.875 17.02  0  1    4    4     Four
#> 3 22.8   4  108  93 3.85 2.320 18.61  1  1    4    1     Four
#> 4 21.4   6  258 110 3.08 3.215 19.44  1  0    3    1    Three
#> 5 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2    Three
#> 6 18.1   6  225 105 2.76 3.460 20.22  1  0    3    1    Three

gear_df %>%
  mutate(gear_fct = fct_relevel(gear_chr, "Three", "Four", "Five")) %>%
  ggplot(aes(x = gear_fct)) +
  geom_bar() +
  labs(x = "Gear", y = "Number of cars")

Exercise 2

forcats::gss_cat

#> # A tibble: 21,483 x 9
#>     year marital     age race  rincome    partyid     relig     denom    tvhours
#>    <int> <fct>     <int> <fct> <fct>      <fct>       <fct>     <fct>      <int>
#>  1  2000 Never ma…    26 White $8000 to … Ind,near r… Protesta… Souther…      12
#>  2  2000 Divorced     48 White $8000 to … Not str re… Protesta… Baptist…      NA
#>  3  2000 Widowed      67 White Not appli… Independent Protesta… No deno…       2
#>  4  2000 Never ma…    39 White Not appli… Ind,near r… Orthodox… Not app…       4
#>  5  2000 Divorced     25 White Not appli… Not str de… None      Not app…       1
#>  6  2000 Married      25 White $20000 - … Strong dem… Protesta… Souther…      NA
#>  7  2000 Never ma…    36 White $25000 or… Not str re… Christian Not app…       3
#>  8  2000 Divorced     44 White $7000 to … Ind,near d… Protesta… Luthera…      NA
#>  9  2000 Married      44 White $25000 or… Not str de… Protesta… Other          0
#> 10  2000 Married      47 White $25000 or… Strong rep… Protesta… Souther…       3
#> # … with 21,473 more rows

relig <- gss_cat %>%
  group_by(relig) %>%
  summarize(tvhours = mean(tvhours, na.rm = TRUE))

#> `summarise()` ungrouping output (override with `.groups` argument)


ggplot(relig, aes(tvhours, relig)) +
  geom_point()

relig <- gss_cat %>%
  group_by(relig) %>%
  summarize(tvhours = mean(tvhours, na.rm = TRUE))

#> `summarise()` ungrouping output (override with `.groups` argument)


relig %>%
  mutate(relig = fct_reorder(relig, tvhours)) %>%
  ggplot(aes(tvhours, relig)) +
  geom_point()

Bonus: Exercise 3

library(ggridges)

ggplot(gss_cat, aes(x = tvhours, y = relig, fill = relig)) +
  geom_density_ridges(alpha = 0.8) +
  labs(x = 'Number of hours spent watching TV', y = 'Religion') +
  guides(fill = FALSE) +
  theme_minimal()

#> Picking joint bandwidth of 0.586

#> Warning: Removed 10146 rows containing non-finite values (stat_density_ridges).