Create a column that assigns value to a row in a dataframe based on an event in another row

Question

I have a dataframe that is structured like the following:

example <- data.frame(id = c(1,1,1,1,1,1,1,2,2,2,2,2),
                      event = c("email","email","email","draw","email","email","draw","email","email","email","email","draw"),
                      date = c("2020-03-01","2020-06-01","2020-07-15","2020-07-28","2020-08-07","2020-09-01","2020-09-15","2020-05-22","2020-06-15","2020-07-13","2020-07-15","2020-07-31"),
                      amount = c(NA,NA,NA,10000,NA,NA,1500,NA,NA,NA,NA,2200))

This is a simplified version of the dataframe. I am trying to create a column that will assign a 1 to the last email before the draw event and a column that will have the amount drawn on the same row as the email. The desired dataframe would look like the following:

desiredResult <- data.frame(id = c(1,1,1,1,1,1,1,2,2,2,2,2),
                      event = c("email","email","email","draw","email","email","draw","email","email","email","email","draw"),
                      date = c("2020-03-01","2020-06-01","2020-07-15","2020-07-28","2020-08-07","2020-09-01","2020-09-15","2020-05-22","2020-06-15","2020-07-13","2020-07-15","2020-07-31"),
                      amount = c(NA,NA,NA,10000,NA,NA,1500,NA,NA,NA,NA,2200),
                      EmailBeforeDrawFlag = c(NA,NA,1,NA,NA,1,NA,NA,NA,NA,1,NA),
                      EmailBeforeDrawAmount = c(NA,NA,10000,NA,NA,1500,NA,NA,NA,NA,2200,NA))

Answer 1

Here is the dplyr solution. When you create new columns, you want to use if_else() in the definition of EmailBeforeDrawFlag to check a condition, and the lead function to look in the previous row for event . EmailBeforeDrawAmount is juts lead(amount) .

example %>%
  mutate(EmailBeforeDrawFlag = if_else(lead(event) == "draw", 1, NA_real_ ),
         EmailBeforeDrawAmount = lead(amount))
   id event       date amount EmailBeforeDrawFlag EmailBeforeDrawAmount
1   1 email 2020-03-01     NA                  NA                    NA
2   1 email 2020-06-01     NA                  NA                    NA
3   1 email 2020-07-15     NA                   1                 10000
4   1  draw 2020-07-28  10000                  NA                    NA
5   1 email 2020-08-07     NA                  NA                    NA
6   1 email 2020-09-01     NA                   1                  1500
7   1  draw 2020-09-15   1500                  NA                    NA
8   2 email 2020-05-22     NA                  NA                    NA
9   2 email 2020-06-15     NA                  NA                    NA
10  2 email 2020-07-13     NA                  NA                    NA
11  2 email 2020-07-15     NA                   1                  2200
12  2  draw 2020-07-31   2200                  NA                    NA

Answer 2

We could also make use of NA^ to create the column on the lead

library(dplyr)
example %>%
      mutate(EmailBeforeDrawFlag = NA^(lead(event != 'draw')), 
             EmailBeforeDrawAmount = lead(amount))

-output

#    id event       date amount EmailBeforeDrawFlag EmailBeforeDrawAmount
#1   1 email 2020-03-01     NA                  NA                    NA
#2   1 email 2020-06-01     NA                  NA                    NA
#3   1 email 2020-07-15     NA                   1                 10000
#4   1  draw 2020-07-28  10000                  NA                    NA
#5   1 email 2020-08-07     NA                  NA                    NA
#6   1 email 2020-09-01     NA                   1                  1500
#7   1  draw 2020-09-15   1500                  NA                    NA
#8   2 email 2020-05-22     NA                  NA                    NA
#9   2 email 2020-06-15     NA                  NA                    NA
#10  2 email 2020-07-13     NA                  NA                    NA
#11  2 email 2020-07-15     NA                   1                  2200
#12  2  draw 2020-07-31   2200                  NA                    NA