Mark all values given value in dataframe

Question

I have a dataframe with time data, with which I'd like to mark certain rows as belonging to a period of interest. For example:

time <- c(1,2,3,5,7,9,23,24,28,43,45)
action <- c("clap","blink","stare","stare","clap","stare",
"clap","blink","stare","clap","stare")

I'd like to find what happens within, say, 5 second of each clap, hopefully by creating a new column which would have an index marker for each new sequence of clap and post-clap.

I'm quite a novice with R so there might be a way of phrasing this that would help with searching for an answer, so my apologies if that is the case.

Answer 1

I believe this does what you want. Note that I have added a row to your data.frame , since you mention you want an identifier to see if something happens within 5 seconds from a clap.

# Your data
time <- c(1,2,3,5,7,9,23,24,28,43,45)
action <- c("clap","blink","stare","stare","clap","stare",
            "clap","blink","stare","clap","stare")
df = data.frame(time=time,action=action)

# Added a row that does not occur within 5 seconds of a clap for illustration purposes.
df = rbind(df,data.frame(time=100,action='stare'))

# Add the group number, based on the claps.
df = df %>% mutate(group=cumsum(action=='clap'))

# Check if action is within 5 seconds of a clap.
df = df %>% 
  group_by(group) %>% 
  mutate(within_5_seconds = ifelse(time-time[1]<=5,1,0)) %>%
  as.data.table

Output:

    time action group within_5_seconds
 1:    1   clap     1                1
 2:    2  blink     1                1
 3:    3  stare     1                1
 4:    5  stare     1                1
 5:    7   clap     2                1
 6:    9  stare     2                1
 7:   23   clap     3                1
 8:   24  blink     3                1
 9:   28  stare     3                1
10:   43   clap     4                1
11:   45  stare     4                1
12:  100  stare     4                0