---

In Excel, pivot tables offer an easy way to group and summarize data.

For example, if we have the following dataset in Excel then we can use a pivot table to quickly summarize the total sales by region:

This tells us:

• Region A had 51 total sales
• Region B had 85 total sales
• Region C had 140 total sales

Or we could summarize by another metric such as the average sales by region:

It turns out that we can quickly create similar pivot tables in R by using the group_by() and summarize() functions from the package.

This tutorial provides several examples of how to do so.

Example: Create Pivot Tables in R

First, let’s create the same dataset in R that we used in the previous examples from Excel:

#create data frame
df <- data.frame(region=c('A', 'A', 'A', 'B', 'B', 'B', 'C', 'C', 'C', 'C'),
                 device=c('X', 'X', 'Y', 'X', 'Y', 'Y', 'X', 'X', 'Y', 'Y'),
                 sales=c(12, 18, 21, 22, 34, 29, 38, 36, 34, 32))

#view data frame
df

   region device sales
1       A      X    12
2       A      X    18
3       A      Y    21
4       B      X    22
5       B      Y    34
6       B      Y    29
7       C      X    38
8       C      X    36
9       C      Y    34
10      C      Y    32

Next, let’s load the dplyr package and use the group_by() and summarize() functions to group by region and find the sum of sales by region:

library(dplyr)#find sum of sales by region
df %>%
  group_by(region) %>% 
  summarize(sum_sales = sum(sales))

# A tibble: 3 x 2
  region sum_sales
        
1 A             51
2 B             85
3 C            140

We can see that these numbers match the numbers shown in the introductory Excel example.

We can also calculate the average sales by region:

#find average sales by region
df %>%
  group_by(region) %>% 
  summarize(mean_sales = mean(sales))

# A tibble: 3 x 2
  region  mean_sales
        
1 A             17  
2 B             28.3
3 C             35  

Note that we can also group by multiple variables. For example, we could find the sum of sales grouped by region and device type:

#find sum of sales by region and device type
df %>%
  group_by(region, device) %>% 
  summarize(sum_sales = sum(sales))

# A tibble: 6 x 3
# Groups:   region [3]
  region device sum_sales
          
1 A      X             30
2 A      Y             21
3 B      X             22
4 B      Y             63
5 C      X             74
6 C      Y             66  

Additional Resources