[R] algorithm help

[Carl Witthoft]

try this:

?rle


Carl



******
From: array chip <arrayprofile_at_yahoo.com>
Date: Thu, 06 Jan 2011 14:16:38 -0800 (PST)


Hi, I am seeking help on designing an algorithm to identify the 
locations of stretches of 1s in a vector of 0s and 1s. Below is an 
simple example:

 > dat<-as.data.frame(cbind(a=c(F,F,T,T,T,T,F,F,T,T,F,T,T,T,T,F,F,F,F,T)
   ,b=c(4,12,13,16,18,20,28,30,34,46,47,49,61,73,77,84,87,90,95,97)))

 > dat

    a b
1 0 4
2 0 12
3 1 13
4 1 16
5 1 18
6 1 20
7 0 28
8 0 30
9 1 34
10 1 46
11 0 47
12 1 49
13 1 61
14 1 73
15 1 77
16 0 84
17 0 87
18 0 90
19 0 95
20 1 97

In this dataset, "b" is sorted and denotes the location for each number 
in "a". So I would like to find the starting & ending locations for each 
stretch of 1s within "a", also counting the number of 1s in each stretch 
as well. Hope the results from the algorithm would be: