CODEWITHSUNDEEP
algorithmrdate
BlueTrin
BlueTrin

Reputation: 10103

In R, equivalent of upper_bound()

I have two vectors of dates A and B. I would like for each date of A, obtain the closest date in B which is greater or superior to the date itself.

(the reason is that B is a list of working days and I need my output to be a working day).

How can I do this in a vectorised manner in R ?

Upvotes: 2

Views: 133

Answers (2)

Andy
Andy

Reputation: 4659

+1 to blindJesse for the apply logic, but care needs to be given to the case where no nearest date exists as well as the the coercion to numeric that occurs:

a <- as.Date(sample(1:20, 5, T), origin=Sys.Date())
#[1] "2012-08-26" "2012-08-31" "2012-08-25" "2012-08-18" "2012-08-20"
b <- as.Date(sample(1:20, 5, T), origin=Sys.Date())
#[1] "2012-08-27" "2012-08-27" "2012-08-25" "2012-08-22" "2012-08-17"

sapply(a, function(x) min(b[b>x]))
#[1] 15579   Inf 15579 15574 15574

# generate the min index instead, catching for no min case
min.indices <- sapply(a, function(x) {
  ifelse(length(which.min(b[b>x]))==0, NA, which.min(b[b>x]))
})

b[min.indices]
#[1] "2012-08-27" NA           "2012-08-27" "2012-08-22" "2012-08-22"

Upvotes: 2

Jesse Anderson
Jesse Anderson

Reputation: 4613

This should work, though there might be a vectorized solution:

sapply(d1, function(x) min(d2[d2>x]))

Upvotes: 3

Related Questions