R help - Jul 2011 - UNIX diff function

Colleagues,

(R: 2.13.0; OS X)

I often receive sequential datasets in which there are new rows interposed
between existing rows.  For example:
	SET1 <- data.frame(list(LETTERS=LETTERS[c(1:4, 6:10)], NUMBERS=c(1:4,
6:10)))
	SET2 <- data.frame(list(LETTERS=LETTERS[1:10], NUMBERS=1:10))
> SET1
  LETTERS NUMBERS
1       A       1
2       B       2
3       C       3
4       D       4
5       F       6
6       G       7
7       H       8
8       I       9
9       J      10
> SET2
   LETTERS NUMBERS
1        A       1
2        B       2
3        C       3
4        D       4
5        E       5
6        F       6
7        G       7
8        H       8
9        I       9
10       J      10

As you can see, the row containing E and 5 was inserted into the second set. 
The UNIX diff command identifies the differences quite readily.  Obviously, the
R diff function does not do this.  However, one kluge that I use is to paste
together all the entries in each row, then perform a setdiff on the two
resulting vectors.  Assuming that no rows are duplicated (which would true in my
data), my approach works but is it cumbersome.

I suspect that someone on this board has thought of a more clever approach to
this (or perhaps some function already exists).  Any help would be appreciated.

Thanks.

Dennis


Dennis Fisher MD
P < (The "P Less Than" Company)
Phone: 1-866-PLessThan (1-866-753-7784)
Fax: 1-866-PLessThan (1-866-753-7784)
PLessThan.com

Hi Dennis,

It still uses paste(), but this isn't so bad:
SET2[!(do.call(paste, SET2) %in% do.call(paste, SET1)),]

You could even turn it into a function. This one checks for matching
rows in the data frame with more rows, but you could take out the
conditional if you want order of arguments to be important.

rowdiff <- function(df1, df2) {
   if(nrow(df1) >= nrow(df2)) {
      df1[!(do.call(paste, df1) %in% do.call(paste, df2)),]
   } else {
      df2[!(do.call(paste, df2) %in% do.call(paste, df1)),]
   }
}
> rowdiff(SET1, SET2)
  LETTERS NUMBERS
5       E       5
> rowdiff(SET2, SET1)
  LETTERS NUMBERS
5       E       5
> rowdiff(SET1, SET1)
[1] LETTERS NUMBERS
<0 rows> (or 0-length row.names)


Sarah

On Wed, Jul 13, 2011 at 1:14 PM, Dennis Fisher <fisher at plessthan.com>
wrote:
> Colleagues,
>
> (R: 2.13.0; OS X)
>
> I often receive sequential datasets in which there are new rows interposed
between existing rows. ?For example:
> ? ? ? ?SET1 <- data.frame(list(LETTERS=LETTERS[c(1:4, 6:10)],
NUMBERS=c(1:4, 6:10)))
> ? ? ? ?SET2 <- data.frame(list(LETTERS=LETTERS[1:10], NUMBERS=1:10))
>
>> SET1
> ?LETTERS NUMBERS
> 1 ? ? ? A ? ? ? 1
> 2 ? ? ? B ? ? ? 2
> 3 ? ? ? C ? ? ? 3
> 4 ? ? ? D ? ? ? 4
> 5 ? ? ? F ? ? ? 6
> 6 ? ? ? G ? ? ? 7
> 7 ? ? ? H ? ? ? 8
> 8 ? ? ? I ? ? ? 9
> 9 ? ? ? J ? ? ?10
>
>> SET2
> ? LETTERS NUMBERS
> 1 ? ? ? ?A ? ? ? 1
> 2 ? ? ? ?B ? ? ? 2
> 3 ? ? ? ?C ? ? ? 3
> 4 ? ? ? ?D ? ? ? 4
> 5 ? ? ? ?E ? ? ? 5
> 6 ? ? ? ?F ? ? ? 6
> 7 ? ? ? ?G ? ? ? 7
> 8 ? ? ? ?H ? ? ? 8
> 9 ? ? ? ?I ? ? ? 9
> 10 ? ? ? J ? ? ?10
>
> As you can see, the row containing E and 5 was inserted into the second
set. ?The UNIX diff command identifies the differences quite readily.
?Obviously, the R diff function does not do this. ?However, one kluge that I use
is to paste together all the entries in each row, then perform a setdiff on the
two resulting vectors. ?Assuming that no rows are duplicated (which would true
in my data), my approach works but is it cumbersome.
>
> I suspect that someone on this board has thought of a more clever approach
to this (or perhaps some function already exists). ?Any help would be
appreciated.
>
> Thanks.
>
> Dennis
>
>
-- 
Sarah Goslee
functionaldiversity.org