[R] Replace missing value within group with non-missing value

David Winsemius dwinsemius at comcast.net
Sat Apr 6 18:26:26 CEST 2013


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:ruipbarradas at sapo.pt] 
> Sent: 06 April 2013 16:32
> To: Leask, Graham
> Cc: r-help at 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 '$<-.data.frame' ('*tmp*', "mth", value = NA_real_) :
>>     replacement has 1 rows, data has 0
>> 
>> I'm sure there must be an elegant solution to this problem?
>> 
>> Best wishes
>> 
>> 
>> 
>> Graham
>> 
>> On 6 Apr 2013, at 12:15, "Rui Barradas" <ruipbarradas at sapo.pt> wrote:
>> 
>>> Hello,
>>> 
>>> That's not a very good way of posting your data, preferably paste the output of ?dput in a post.
>>> Some thing along the lines of the following might do what you want. 
>>> It seems that the groups are established by 'dn' and 'obs' numbers. 
>>> If so, try
>>> 
>>> 
>>> # Make up some data
>>> dat <- data.frame(dn = 4, obs = rep(1:5, each = 6), mth = NA) 
>>> dat$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 at 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 at 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



More information about the R-help mailing list