Question

This is a very specific task but I believe the learning from it may be general enough to be of use. I have an oddly designed data set of time stamps that I need to reshape and I need to operate out of the packages that come with a base install. I think seeing is easier than explaining.

The list of named vectors with time stamps:

x <- structure(list(A = c("2.40", ":", "3.00", "5.01", "6.62", ":", 
    "7.00", "9.00"), B = c("2.40", "5.01", "6.62", ":", "7.00", "9.00"
    ), C = c("2.40", ":", "3.00", "5.01", "6.62", ":", "7.00", "9.00"
    )), .Names = c("A", "B", "C"))

What the list looks like:

> x
[[1]]
[1] "2.40" ":"    "3.00" "5.01" "6.62" ":"    "7.00" "9.00"

[[2]]
[1] "2.40" "5.01" "6.62" ":"    "7.00" "9.00"

[[3]]
[1] "2.40" ":"    "3.00" "5.01" "6.62" ":"    "7.00" "9.00"

What I’d like:

$A
     start      end  
1   "2.40"   "3.00"   
2   "5.01"   "5.01"    
3   "6.62"   "7.00"   
4   "9.00"   "9.00"  

$B
     start      end 
1   "2.40"   "2.40"   
2   "5.01"   "5.01"  
3   "6.62"   "7:00"   
4   "9.00"   "9.00"   

$C
     start      end 
1   "2.40"   "3.00"
2   "5.01"   "5.01" 
3   "6.62"   "7.00"
4   "9.00"   "9.00"

Where there’s a colon (:) the element on the left is a start value and the element on the right is an end value. If an element is not touching a colon it needs to be repeated and is both the start and end value.

Note: the outcome wouldn’t have quotes if it’s a dataframe vs. a matrix

Answer 1

Is this what you want?

> lapply(x, function(v)
+   data.frame(start = v[-c(f <- which(v==":"), f + 1)], 
+              end = v[-c(f, f-1)]))
$A
  start  end
1  2.40 3.00
2  5.01 5.01
3  6.62 7.00
4  9.00 9.00

$B
  start  end
1  2.40 2.40
2  5.01 5.01
3  6.62 7.00
4  9.00 9.00

$C
  start  end
1  2.40 3.00
2  5.01 5.01
3  6.62 7.00
4  9.00 9.00