Summarizing Data with Tidyverse

Summarizing data

Duration: ~20 Minutes

Learning Objectives

Summarize data
Group and ungroup variables

Why learn `Summarize()`

Base R has many ways to summarize data.

With summarize(), get code that is

Cleaner
More readable
Shorter (sometimes)

Example: column means

Tidyverse is sometimes longer,
but always clearer.

Base R


colMeans(iris[, 1:2])

# or more explicitly,
colMeans(
  iris[, c("Sepal.Length",
           "Sepal.Width")] )

# Sepal.Length  Sepal.Width 
#     5.843333     3.057333

Tidyverse


iris %>% summarize(
    mean_slength = 
      mean(Sepal.Length),
    mean_swidth = 
      mean(Sepal.Width)
  )
  
#   mean_slength mean_swidth
# 1     5.843333    3.057333

Example: category means

Base R


data.frame(mean_petal_length = 
  c(setosa = mean(
      iris$Petal.Length
      [iris$Species=="setosa"]
    ), versicolor = mean(
      iris$Petal.Length
      [iris$Species=="versicolor"]
    ), virginica = mean(
      iris$Petal.Length
      [iris$Species=="virginica"]
    )))
    
#            mean_petal_length
# setosa                 1.462
# versicolor             4.260
# virginica              5.552

Tidyverse


iris %>% group_by(Species) %>%
  summarize(
    mean_petal_length = 
      mean(Petal.Length)
  ) %>% ungroup()





#   Species    mean_petal_length
# 1 setosa                  1.46
# 2 versicolor              4.26
# 3 virginica               5.55

Your Turn: `summarize()`

Calculate the sums of Petal.Length and Petal.Width.
Save your results to a dataset.


my_iris <- iris %>% summarize(
  sum_petal_length = sum(Petal.Length),
  sum_petal_width = sum(Petal.Width)
)

Your Turn: `group_by()`

Calculate the sums and of Petal.Length and Petal.Width grouped by Species.
Save your results to a dataset.


my_iris <- iris %>% group_by(Species) %>%
  summarize(
    sum_petal_length = sum(Petal.Length),
    sum_petal_width = sum(Petal.Width)
  ) %>% ungroup()

Summarize rows

Summarize across rows instead of columns.

Base R


rowMeans(iris[, 1:4])



# [1] 2.550 2.375 2.350 2.350 
# [5] 2.550 2.850 2.425 2.525 
# [9] 2.225 2.400 ...

Tidyverse


iris %>% 
  select(Sepal.Length, 
         Sepal.Width, 
         Petal.Length, 
         Petal.Width) %>%
  rowMeans()

But what if we are calculating something more complex than means?

Bonus: `rowwise()`

In the code below, replace mean with your function of choice.


iris %>% rowwise() %>%
  mutate(
    mean = mean(c(Sepal.Length, 
                  Sepal.Width, 
                  Petal.Length, 
                  Petal.Width))
  )

What happens if you leave out rowwise()?

Another example

Now we are switching to cars.

Base R


## Easy, but limited
c(n_rows = nrow(cars), 
  colMeans(cars))


n_rows  speed   dist 
 50.00  15.40  42.98

dplyr


## More complex, more powerful
cars %>%
  summarize(n_rows = n(),
            speed = mean(speed),
            dist = mean(dist) 
  )

  
  n_rows  speed   dist
1     50   15.4  42.98

And Now You Know!

Next up: Summarize practice