CODEWITHSUNDEEP
rdplyr
user3483060
user3483060

Reputation: 407

generate new column with cumulative sum using dplyr

I am trying to do the next exercise:

data<-tbl_df(data.frame(A=c(6,3,2,4,5,3,3,4),
                        B=c(10,19,12,10,10,9,8,20),
                        C=c("A","B","C","A","A","C","C","B"),
                        D=c("AB","AB","AB","CD","CD","AB","CD","CD")))

data%>%group_by(C,D)%>%
  summarise(Frecuencia=n(),suma=sum(A),cumsum(suma))

I am trying to add a new column with the cumulative sum of the column suma, but It doesn´t works.

    > data%>%group_by(C,D)%>%
+   summarise(Frecuencia=n(),suma=sum(A),cumsum(suma))
# A tibble: 6 x 5
# Groups:   C [?]
  C     D     Frecuencia  suma `cumsum(suma)`
  <fct> <fct>      <int> <dbl>          <dbl>
1 A     AB             1     6              6
2 A     CD             2     9              9
3 B     AB             1     3              3
4 B     CD             1     4              4
5 C     AB             2     5              5
6 C     CD             1     3              3

Why is not it working?

Trying with mutate in this example works, but in my real data I can´t get the result.

Upvotes: 1

Views: 2080

Answers (1)

Naveen
Naveen

Reputation: 1210

I think you cannot create a new column with a column that is being created by a function. Alternatively, you can use mutate function and add a new column keeping the existing columns.

This code cumsums based on each value of column C. If you can provide the desired output, I can answer your questions more accurate.

Code:

data %>% group_by(C,D) %>%
  summarise(Frecuencia=n(),suma=sum(A)) %>%
  mutate(cumsum(suma))

Output:

enter image description here

Upvotes: 2

Related Questions