Writing a function that restarts cumulative sum whenever a repetition of entries is broken

Question

I am attempting to write a function in R that is similar to cumsum() but only adds up entries that are the same, and restarts the count once the entry changes. So, my data includes a vector of two levels: 0 or 1. I would like to count up the number of entries that all say 1 in a row, and when we get to a 0 the counter should restart back to 1. Hopefully that makes sense, I can clarify if need be.

So I have this fake data:

df <- data.frame(repeats = c(1, 0, 0 , 0, 1, 1, 1, 1, 0, 0, 1, 0, 1, 1))

I expect my output to be:

1, 3, 4, 2, 1, 1, 2

But I am having much difficulty. I know my code is very wrong. Here is what I have:

cumsum2 <- function(x) {
    cumulative <- rep(0, 19)
    
    for (i in cumulative) {
      
      if (i == 1) {
        
        cumulative[i] <- x[[1]] # set first entry as value of x[1]
        
      } else if (i > 1) {
        
        if (x[i] == x[i-1]) { # set all other entries as cumulative sum
          
          sum <- sum(x[i] + x[i-1])  # as long as they match previous entry
          
          cumulative[i] <- sum
          
        } else { # if they don't match previous entry, restart cumsum
          
          cumulative[i] <- x[[i]]
          
        }
        
      }
      
    }
    
    return(cumulative)
}

and my current output is unchanged:

cumsum2(df$repeats)

[1] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Could someone please point me in the right direction? I have not coded for quite some time and I feel like I have forgotten so much, but I am hoping it will start to come back quickly. Thank you in advance.

Answer 1

You can use rle

rle(df$repeats)$lengths

Output:

[1] 1 3 4 2 1 1 2