在多年事件的R表中，如何标记另一个事件的首次出现？

Question

I'm working with an R table right now that stores events that take place over multiple years. Each of these events has several related variables. For one of these variable, I want to pull out the first occurrence of it in each event series.

events <- c('event1', 'event1', 'event1', 'event1', 'event1', 'event2', 'event2', 'event2')  
years <- c('2000', '2001', '2002', '2003', '2004', '1994', '1995', '1996') 
variable1 <- c('False', 'False', 'False', 'True', 'True', 'False', 'False', 'True')  
df <- data.frame(events, years, variable1)  

I want to figure out a way to generate a new column, First_occurrence, that looks like this:

Event   Year   Variable1   First_occurrence
event1  1994   False       False                              
event1  1995   True        True
event1  1996   True        False
event1  1997   True        False
event2  2000   False       False
event2  2001   False       False
event2  2002   True        True

How would I go about creating that "First_occurrence" column?

Thanks!

Answer 1

Something like this, if you're ok with tidyverse solutions:

library(dplyr)
df %>% 
 arrange(events,years) %>% 
 group_by(events) %>% 
 mutate(first_occ = row_number() == which(variable1 == 'True')[1])

# A tibble: 8 x 4
# Groups:   events [2]
  events years variable1 first_occ
   <chr> <chr>     <chr>     <lgl>
1 event1  2000     False     FALSE
2 event1  2001     False     FALSE
3 event1  2002     False     FALSE
4 event1  2003      True      TRUE
5 event1  2004      True     FALSE
6 event2  1994     False     FALSE
7 event2  1995     False     FALSE
8 event2  1996      True      TRUE

A small note: stuff like this could potentially be slightly smoother to code if your "True"/"False" variables are actually booleans, rather than character or factors. In that case boolean comparisons are easier and you can leverage the automatic coercion to 1/0 in many cases.

Note that this answer may yield undesirable results if an event has no occurrences at all.