R help - Dec 2009 - documentation of intersect() on string vector and num vector and on duplicated elements

> intersect(c(1,3,2),c('1','3'))
[1] "1" "3"

Apparently, intersect() treats num as string. But this is not
documented in the help. Could somebody add it in the future version of
R?

Also according to the help, the argument should not have duplicated
elements. But I tried the following example, it seems that it doesn't
matter where there are duplicated elements or not. Should the help be
corrected? Or somebody could provide me with an example that make
intersect() fails if there are duplicated elements?
> x=c(3,1,2,1,7)
> y=c(5,1,4,2,3,8,1)
> x
[1] 3 1 2 1 7
> y
[1] 5 1 4 2 3 8 1
> intersect(x,y)
[1] 3 1 2

> > intersect(c(1,3,2),c('1','3'))
> [1] "1" "3"
> 
> Apparently, intersect() treats num as string. But this is not
> documented in the help. Could somebody add it in the future version of
> R?
> 
This is more a general coercion feature. For example 1 == "1", see
?Comparison.
> Also according to the help, the argument should not have duplicated
> elements. 
I think this is just because these functions were written with sets in mind,
which would not have duplicated values, thus the word "conceptually". 
The duplicated elements are simply discarded.

On Dec 2, 2009, at 4:33 PM, Peng Yu wrote:
>> intersect(c(1,3,2),c('1','3'))   # note x is numeric
and y is
>> character
> [1] "1" "3"
>
> Apparently, intersect() treats num as string. But this is not
> documented in the help. Could somebody add it in the future version of
> R?
It is documented that intersect will return a value of that is same  
mode as its "y" argument. How could it be any more clear?
>
> Also according to the help, the argument should not have duplicated
> elements.
That's not what my help page says. It says that "conceptually"
they
won't have them, but immediately thereafter it says:
"Details
Each of union, intersect, setdiff and setequal will discard any  
duplicated values in the arguments, and they apply as.vector to their  
arguments (and so in particular coerce factors to character vectors)."

Unfortunately, R does not have a package that will ensure that users  
will read the help pages carefully.

-- 

David
> But I tried the following example, it seems that it doesn't
> matter where there are duplicated elements or not. Should the help be
> corrected? Or somebody could provide me with an example that make
> intersect() fails if there are duplicated elements?
>
>> x=c(3,1,2,1,7)
>> y=c(5,1,4,2,3,8,1)
>> x
> [1] 3 1 2 1 7
>> y
> [1] 5 1 4 2 3 8 1
>> intersect(x,y)
> [1] 3 1 2
>
> ______________________________________________
> 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, MD
Heritage Laboratories
West Hartford, CT

> 
> It is documented that intersect will return a value of that is same
> mode as its "y" argument. How could it be any more clear?
> 
Oops, this makes my previous response to this question inaccurate, apologies.