comparing columns in a dataframe

d <- data.frame(x=1:10,y=sin(1:10),z=factor(letters[1:10]))
d1 <- d
d1$x[2:3] <- jitter(d$x[2:3] )
d2 <- subset(d1, !(z %in% c("a","g")))

compare(d,d2,allowAll=T)->test
str(test)

  Hi: you've got to create a setdiff in both directions in order to  
get the
  lone ones in each column because setdiff is
  not? commutative meaning that setdiff(a,b) does not equal  
setdiff(b,a). once
  you do that, then
  ( setdiff1 + setdiff2 - intersect )? should equal the? union.
  if it doesn't, that would be weird and more investigation would  
need to be
  done.

  On Apr 4, 2009, Bob Green <bgreen at dyson.brisnet.org.au> wrote:

    hello,
    I am hoping for some advice regarding comparing variables from 3
    versions of a spreadsheet which have been combined into a single
    dataframe. The aim is to identify which rows have been changed.
    The dataframe contains 177 rows of data (each cell contains text).
    'intersect' produced a file with 35 rows, 'union' a file with 303
    rows and 'setdiff' a file with 130 rows
    Below is the code that I have started with.
    Ideally I would like to identify the actual row numbers where  
there
    is difference in the variables (either pairwise or between 3  
variables).
    x <- read.csv("c://rec_compare.csv",header=T, as.is=TRUE)
    u <- union(x$rm1, x$redc1)
    write.csv(u,"c:/union_test.csv")
    i <- intersect(x$rm1, x$redc1)
    write.csv(i,"c:/intersect_test.csv")
    sd <- setdiff(x$rm1, x$redc1)
    write.csv(sd,"c:/setdiff_test.csv")
    Any suggestions are appreciated.
    regards
    Bob