On Apr 6, 2013, at 9:16 AM, Leask, Graham wrote: > Hi Rui, > > Data as follows > > structure(list(dn = c(4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, > 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, > 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4), obs = c(1, 1, > 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, > 4, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, 7, 7, 7, 7, 7, 7, 8, 8, > 8, 8, 8, 8, 9, 9), choice = c(0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, > 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, > 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0), br = c(1, > 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, > 5, 6, 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6, 1, > 2, 3, 4, 5, 6, 1, 2), mth = c(NA, NA, NA, NA, NA, 487, NA, NA, > 488, NA, NA, NA, NA, NA, NA, NA, NA, 488, NA, NA, 489, NA, NA, > NA, NA, NA, NA, NA, NA, 489, NA, NA, NA, NA, NA, 489, NA, NA, > NA, NA, NA, 490, NA, NA, NA, NA, NA, 491, NA, NA)), .Names = c("dn", > "obs", "choice", "br", "mth"), row.names = c("1", "2", "3", "4", > "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", > "16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", > "27", "28", "29", "30", "31", "32", "33", "34", "35", "36", "37", > "38", "39", "40", "41", "42", "43", "44", "45", "46", "47", "48", > "49", "50"), class = "data.frame") >
Looks like a job for na.locf in the zoo package: require(zoo) # will fail if first value is NA so either this ... dat$mth[-(1:5)] <- na.locf(dat$mth[-(1:5)]) # ... or this: dat$mth <- na.locf(dat$mth, na.rm=FALSE) -- David. > Best wishes > > > Graham > > -----Original Message----- > From: Rui Barradas [mailto:ruipbarra...@sapo.pt] > Sent: 06 April 2013 16:32 > To: Leask, Graham > Cc: r-help@r-project.org > Subject: Re: [R] Replace missing value within group with non-missing value > > Hello, > > Can't you post a data example? If your dataset is named 'dat' use > > dput(head(dat, 50)) # paste the output of this in a post > > > Rui Barradas > > Em 06-04-2013 15:34, Leask, Graham escreveu: >> Hi Rui, >> >> Thank you for your suggestion which is very much appreciated. Unfortunately >> running this code produces the following error. >> >> error in '$<-.data.frame' ('*tmp*', "mth", value = NA_real_) : >> replacement has 1 rows, data has 0 >> >> I'm sure there must be an elegant solution to this problem? >> >> Best wishes >> >> >> >> Graham >> >> On 6 Apr 2013, at 12:15, "Rui Barradas" <ruipbarra...@sapo.pt> wrote: >> >>> Hello, >>> >>> That's not a very good way of posting your data, preferably paste the >>> output of ?dput in a post. >>> Some thing along the lines of the following might do what you want. >>> It seems that the groups are established by 'dn' and 'obs' numbers. >>> If so, try >>> >>> >>> # Make up some data >>> dat <- data.frame(dn = 4, obs = rep(1:5, each = 6), mth = NA) >>> dat$mth[6] <- 487 dat$mth[9] <- 488 dat$mth[18] <- 488 dat$mth[21] <- >>> 489 dat$mth[30] <- 489 >>> >>> >>> sp <- split(dat, list(dat$dn, dat$obs)) >>> names(sp) <- NULL >>> tmp <- lapply(sp, function(x){ >>> idx <- which(!is.na(x$mth))[1] >>> x$mth <- x$mth[idx] >>> x >>> }) >>> do.call(rbind, tmp) >>> >>> >>> Hope this helps, >>> >>> Rui Barradas >>> >>> >>> Em 06-04-2013 11:33, Leask, Graham escreveu: >>>> Dear List members >>>> >>>> I have a large dataset organised in choice groups see sample below >>>> >>>> >>>> +-------------------------------------------------------------------------------------------------+ >>>> | dn obs choice acid br date cdate >>>> situat~n mth year set | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 1. | 4 1 0 LOSEC 1 . . >>>> . . 1 | >>>> 2. | 4 1 0 NEXIUM 2 . . >>>> . . 1 | >>>> 3. | 4 1 0 PARIET 3 . . >>>> . . 1 | >>>> 4. | 4 1 0 PROTIUM 4 . . >>>> . . 1 | >>>> 5. | 4 1 0 ZANTAC 5 . . >>>> . . 1 | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 6. | 4 1 1 ZOTON 6 23aug2000 01:00:00 23aug2000 >>>> NS 487 2000 1 | >>>> 7. | 4 2 0 LOSEC 1 . . >>>> . . 2 | >>>> 8. | 4 2 0 NEXIUM 2 . . >>>> . . 2 | >>>> 9. | 4 2 1 PARIET 3 25sep2000 01:00:00 25sep2000 >>>> L 488 2000 2 | >>>> 10. | 4 2 0 PROTIUM 4 . . >>>> . . 2 | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 11. | 4 2 0 ZANTAC 5 . . >>>> . . 2 | >>>> 12. | 4 2 0 ZOTON 6 . . >>>> . . 2 | >>>> 13. | 4 3 0 LOSEC 1 . . >>>> . . 3 | >>>> 14. | 4 3 0 NEXIUM 2 . . >>>> . . 3 | >>>> 15. | 4 3 0 PARIET 3 . . >>>> . . 3 | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 16. | 4 3 0 PROTIUM 4 . . >>>> . . 3 | >>>> 17. | 4 3 0 ZANTAC 5 . . >>>> . . 3 | >>>> 18. | 4 3 1 ZOTON 6 20sep2000 00:00:00 20sep2000 >>>> R 488 2000 3 | >>>> 19. | 4 4 0 LOSEC 1 . . >>>> . . 4 | >>>> 20. | 4 4 0 NEXIUM 2 . . >>>> . . 4 | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 21. | 4 4 1 PARIET 3 27oct2000 00:00:00 27oct2000 >>>> NL 489 2000 4 | >>>> 22. | 4 4 0 PROTIUM 4 . . >>>> . . 4 | >>>> 23. | 4 4 0 ZANTAC 5 . . >>>> . . 4 | >>>> 24. | 4 4 0 ZOTON 6 . . >>>> . . 4 | >>>> 25. | 4 5 0 LOSEC 1 . . >>>> . . 5 | >>>> >>>> |-------------------------------------------------------------------------------------------------| >>>> 26. | 4 5 0 NEXIUM 2 . . >>>> . . 5 | >>>> 27. | 4 5 0 PARIET 3 . . >>>> . . 5 | >>>> 28. | 4 5 0 PROTIUM 4 . . >>>> . . 5 | >>>> 29. | 4 5 0 ZANTAC 5 . . >>>> . . 5 | >>>> 30. | 4 5 1 ZOTON 6 23oct2000 03:00:00 23oct2000 >>>> NS 489 2000 5 | >>>> >>>> I wish to fill in the missing values in each choice set - delineated by dn >>>> (Doctor) obs (Observation number) and choices (1 to 6). >>>> For each choice set one choice is chosen which contains full time >>>> information for that choice set ie in set 1 choice 6 was chosen and shows >>>> the month 487. The other 5 choices show mth as missing. I want to fill >>>> these with the correct mth. >>>> >>>> I am sure there must be an elegant way to do this in R? >>>> >>>> >>>> Best wishes >>>> >>>> >>>> >>>> Graham >>>> >>>> >>>> [[alternative HTML version deleted]] >>>> >>>> ______________________________________________ >>>> R-help@r-project.org mailing list >>>> https://stat.ethz.ch/mailman/listinfo/r-help >>>> PLEASE do read the posting guide >>>> http://www.R-project.org/posting-guide.html >>>> and provide commented, minimal, self-contained, reproducible code. >>> >> > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. David Winsemius Alameda, CA, USA ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.