How to group by before using count() from the dplyr package-CodePudding

Original Data Frame

route   user_type 
A       member  
A       casual 
A       member
A       member 
A       member  
B       member  
B       casual 
B       member
C       member
C       member

Required new data frame

route  user_type  count_user_type
A      member     4
A      casual     1
B      member     2
B      casual     1
C      member     2

I want to make a new data frame so that I can make a bar plot with route name on Y axis and count_user_type (filled by user_type) on X Axis, I used count(dataframe, route) from plyr package but it gave me the following result

route frequency
A     5    
B     3
C     2

CodePudding user response：

You need to group_by your relevant columns (i.e. to make specific combinations of them to count) before doing a count.

library(dplyr)

df %>% group_by(route, user_type) %>% count(name = "count_user_type")

# A tibble: 5 × 3
# Groups:   route, user_type [5]
  route user_type count_user_type
  <chr> <chr>               <int>
1 A     casual                  1
2 A     member                  4
3 B     casual                  1
4 B     member                  2
5 C     member                  2

CodePudding user response：

library(dplyr)    
df %>% group_by(route, user_type) %>% 
      summarise(count_user_type= n())

CodePudding user response：

You can use the following code:

library(dplyr)
df %>%
  group_by(route, user_type) %>%
  summarise(count_user_type = n())

Output:

# A tibble: 5 × 3
# Groups:   route [3]
  route user_type count_user_type
  <chr> <chr>               <int>
1 A     causal                  1
2 A     member                  4
3 B     causal                  1
4 B     member                  2
5 C     member                  2

Data

df <- data.frame(route = c("A", "A", "A", "A", "A", "B", "B", "B", "C", "C"),
                 user_type = c("member", "causal", "member", "member", "member", "member", "causal", "member", "member", "member"))