Question

i would like to categorize a column in a dataframe based on a comparison of subsequent rows.

for:

DF <- data.frame(respondent=rep(letters[1:2], each=5), response=c(1,1,2,2,1,3,1,1,1,1))
   respondent response
1           a        1
2           a        1
3           a        2
4           a        2
5           a        1
6           b        3
7           b        1
8           b        1
9           b        1
10          b        1

I would like to add a new column (eg: check) that turns to 1 if response in row1 = response in row2 and 0 in case they are not the same. This should be done separately for each respondent.

This would give me

   respondent response check
1           a        1  1
2           a        1  0
3           a        2  1
4           a        2  0
5           a        1  
6           b        3  0
7           b        1  1
8           b        1  1
9           b        1  1
10          b        1

I think i can figure this out with a foor loop but it seems to be a suitable problem for ddply … I just do not see how to address comparisons over adjacent rows …

Answer 1

This is a good candidate for plyr since it splits a data and applies a function then returns the data. In this case, you need to consider the whole response vector and a similar vector shifted by one.

The way I have approached this problem in the past is:

shift.func <- function(df) {
  lag<-c(df$response[-1],0)
  df$check <- ifelse(df$response==lag,1,0)

  return(df)
}

ddply(DF,.(respondent),shift.func)