[R] Odp: applying ifelse to dataframe

[Petr PIKAL]

Hi

r-help-bounces at r-project.org napsal dne 22.06.2010 08:28:04:

> The following dataframe will illustrate the problem
> 
> 
DF<-data.frame(name=rep(1:5,each=2),x1=rep("A",10),x2=seq(10,19,by=1),x3=rep
> (NA,10),x4=seq(20,29,by=1))
>  DF$x3[5]<-50
> 
>  # we have a data frame. we are interested in the columns x2,x3,x4 which
> contain sparse
>  # values and many NA.
>  DF
>    name x1 x2 x3 x4
> 1     1  A 10 NA 20
> 2     1  A 11 NA 21
> 3     2  A 12 NA 22
> 4     2  A 13 NA 23
> 5     3  A 14 50 24
> 6     3  A 15 NA 25
> 7     4  A 16 NA 26
> 8     4  A 17 NA 27
> 9     5  A 18 NA 28
> 10    5  A 19 NA 29
> 
> # we have a list of "target values that we want to search for in the 
data
> frame
> # if the value is in the data frame we want to keep it there, otherwise,
>  replace it with NA
> 
> targets<-c(11,12,13,16,19,50,27,24,22,26)
> # so we apply a test by column to the last 3 columns using the "in" test
> # this gives us a mask of whether the data frame 'contains' elements in 
the
> # target list
> 
> mask<-apply(DF[,3:5],2, "%in%" ,targets)
> mask
> 
>          x2    x3    x4
>  [1,] FALSE FALSE FALSE
>  [2,]  TRUE FALSE FALSE
>  [3,]  TRUE FALSE  TRUE
>  [4,]  TRUE FALSE FALSE
>  [5,] FALSE  TRUE  TRUE
>  [6,] FALSE FALSE FALSE
>  [7,]  TRUE FALSE  TRUE
>  [8,] FALSE FALSE  TRUE
>  [9,] FALSE FALSE FALSE
> [10,]  TRUE FALSE FALSE
> 
> # and so DF[2,3] is equal to 11 and 11 is in the target list, so the 
mask is
> True
> # now something like DF<- ifelse(mask==T,DF,NA) is CONCEPTUALLY what I 
want

Data frames are quite clever in preserving their dimensions. I would do

mask=data.frame(a=TRUE, b=TRUE, !mask)

to add column 1 and 2

and

DF[mask]<-NA

Regards
Petr


> to do
> in the end I'd  Like a result that looks like
> 
>    name x1 x2 x3 x4
> 1     1  A NA NA NA
> 2     1  A 11 NA NA
> 3     2  A 12 NA 22
> 4     2  A 13 NANA
> 5     3  A NA 50 24
> 6     3  A NA NA NA
> 7     4  A 16 NA 26
> 8     4  A NA NA 27
> 9     5  A NA NA NA
> 10    5  A 19 NA NA
> 
> Ive tried forcing the DF and the mask into vectors so that ifelse() 
would
> work
> and have tried "apply" using ifelse.. without much luck. any thoughts?
> 
>    [[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.