How do you apply if else statements across user id #? - r

Question

I am trying to create a dummy variable that flags the user id of people who attended a specific event. Each user id has multiple rows and I would like this dummy variable to apply to every row of the flagged user id. For example, using the data set below, I would like to flag the user IDs of everyone who attended "event b" (using a "1" for attended event b and "0" for did not attend event b). The tricky part is that I want the 1 to appear in every row that matches the user IDs of the people who attended "event b".
I want to use this dummy variable to eventually subset the data so that I can assess the event attending patterns of the users who attended a particular event.
df<-data.frame(id=(100,100,100,101,101,102,102,103,103,103,103),
event=("a","b","c","b","d","a","c","a","c","d","e"))

Answer 1

Consider ifelse and ave, iterating across unique values or levels of event
for(ev in unique(df$event)) { # for(ev in levels(df$event)) {
df[[paste0("event_", ev, "_flag")]] <- with(df, ave(ifelse(event == ev, 1, 0), id, FUN=max))
}
df
# id event event_a_flag event_b_flag event_c_flag event_d_flag event_e_flag
# 1 100 a 1 1 1 0 0
# 2 100 b 1 1 1 0 0
# 3 100 c 1 1 1 0 0
# 4 101 b 0 1 0 1 0
# 5 101 d 0 1 0 1 0
# 6 102 a 1 0 1 0 0
# 7 102 c 1 0 1 0 0
# 8 103 a 1 0 1 1 1
# 9 103 c 1 0 1 1 1
# 10 103 d 1 0 1 1 1
# 11 103 e 1 0 1 1 1

Answer 2

As I understood you want to one-hot encode.
You can use the following code with the dummyVars function of the caret package. Afterwards you aggregate the duplicate rows with the corresponding dplyr function.
library(caret)
library(dplyr)
df<-data.frame(id=c(100,100,100,101,101,102,102,103,103,103,103),
event=c("a","b","c","b","d","a","c","a","c","d","e"))
dmy <- dummyVars(" ~ .", data = df)
trsf <- data.frame(predict(dmy, newdata = df))
aggregate(.~id, trsf, FUN=sum)
id event.a event.b event.c event.d event.e
1 100 1 1 1 0 0
2 101 0 1 0 1 0
3 102 1 0 1 0 0
4 103 1 0 1 1 1

Answer 3

Perhaps I'm using a way to simple approach. Using dplyr and tidyr:
df %>%
mutate(value=1) %>%
pivot_wider(names_from="event", values_fill=0)
returns
# A tibble: 4 x 6
id a b c d e
<dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1 100 1 1 1 0 0
2 101 0 1 0 1 0
3 102 1 0 1 0 0
4 103 1 0 1 1 1