Finding ranges in runs of numbers

Question

I'm trying to find runs of years in a data frame (ideally using plyr)

I'd like to get from this:

require(plyr)

dat<-data.frame(
  name=c(rep("A", 11), rep("B", 11)),
  year=c(2000:2010, 2000:2005, 2007:2011)
  )

To this:

out<-data.frame(
  name=c("A", "B", "B"),
  range=c("2000-2010", "2000-2005", "2007-2011"))

It's easy enough to identify whether each group has a continuous run of years:

ddply(dat, .(name), summarise,
      continuous=(max(year)-min(year))+1==length(year))

How do I go about breaking down group "B" into two ranges?

Any ideas or strategies would be really appreciated.

Thanks

Answer 1

Whether you use a function from "plyr" or from base R, you need to first establish some groups. One way to detect the change in groups since your years are sequential is to look for where diff is not equal to 1. diff creates a vector of length one less than the input vector, so we'll initialize that with "1" and take the cumsum of the result.

Putting that mouthful of an explanation into practice, you can try something like this:

dat$id2 <- cumsum(c(1, diff(dat$year) != 1))

From here, you can use aggregate or your favorite grouping function to get the output you're looking for.

aggregate(year ~ name + id2, dat, function(x) paste(min(x), max(x), sep = "-"))
#   name id2      year
# 1    A   1 2000-2010
# 2    B   2 2000-2005
# 3    B   3 2007-2011

To use range with aggregate, you need to change sep to collapse, as below:

aggregate(year ~ name + id2, dat, function(x) paste(range(x), collapse = "-"))