Reputation: 135
I have a boolean vector in which I want to count the number of occurrences of some patterns.
For example, for the pattern "(1,1)"
and the vector "(1,1,1,0,1,1,1)"
, the answer should be 4.
The only built-in function I found to help is grepRaw
, which finds the occurrences of a particular string in a longer string. However, it seems to fail when the sub-strings matching the pattern overlap:
length(grepRaw("11","1110111",all=TRUE))
# [1] 2
Do you have any ideas to obtain the right answer in this case?
Edit 1
I'm afraid that Rich's answer works for the particular example I posted, but fails in a more general setting:
> sum(duplicated(rbind(c(FALSE,FALSE),embed(c(TRUE,TRUE,TRUE,FALSE,TRUE,TRUE,TRUE),2))))
[1] 3
In this other example, the expected answer would be 0.
Upvotes: 1
Views: 236
Reputation: 2541
Using the function rollapply
you can apply a moving window of width = 2
summing the values. Then you can sum the records where the result is equal to 2 i.e. sum(c(1,1))
library(zoo)
z <- c(1,1,1,0,1,1,1)
sum(rollapply(z, 2, sum) == 2)
Upvotes: 1