Re: [R] The "Sets" package and element equality test

[David Meyer]

By going through the code, I came to a similar conclusion also: it 
seems the match.2.list function must be modified in the following way 
to make it work:

match.2.list <- function(l1,l2){
if (class(l1)=="element"){
 l1 <- l1[[1]]
 l2 <- l2[[1]]
}
 length(l1)==length(l2)
}

because depending whether it is called from the constructor or from 
the %e% function, the assumptions about the input types are not the same.
No, this is not not the problem, it's really the hashing code which 
makes this fail. If you turn hashing off, then the original match.2.list 
function will work as expected:

> sets_options("hash", FALSE)
> match.2.list <- function(l1,l2){
+  length(l1)==length(l2)
+ }
> s <- set(list(1,2),list(3,4))
> lset <- cset(s,matchfun = matchfun(match.2.list))
> lset
{<<list(2)>>}

> list(1,8) %e% lset
[1] TRUE


And also, if instead of working with "standard" objects, I want to 
work with S4 / refClass objects, I cannot build cset objects, because 
I have to go first with the set construction which prompts an error:

Currently, S4 objects are not supported at all as set elements. I will 
have a look on this.

Best
David


Error in as.vector(x, "character") :
  cannot coerce type 'environment' to vector of type 'character'


So far, I don't know how to work around this latter issue.

Thanks again for the package and your help.

Regards

Johnny




On Sat, Oct 8, 2011 at 2:40 PM, David Meyer <mey...@technikum-wien.at 
<mailto:mey...@technikum-wien.at>> wrote:

    Dear Johnny,

    this is a bug in the hashing-code of sets. Use:


    sets_options("hash", FALSE)
    lset <- cset(s, matchfun = matchfun(match.2.list))

    which will work.

    Thanks for pointing this out!

    David

    ----------------

    Hi,

    I tried to use the sets package yesterday, which seems to be a
    very powerful
    package: thanks to the developers. I wanted to define sets of
    lists where 2
    lists are considered equal if they have the same length.
    So, I implemented:

    match.2.list <- function(l1,l2){
     length(l1)==length(l2)
    }

    and then defined my cset as:

    s <- set(list(1,2),list(3,4))
    lset <- cset(s,matchfun(match.2.list))

    so if I now do:
    y <- list(3,4)
    y %e% lset

    I get the correct answer, which is TRUE.

    But if I do:
    x <- list(1,8)
    x %e% lset

    I now get FALSE, even though x is a list of length 2, and should
    thus match
    any of the 2 lists in lset.

    I must be doing something wrong; I checked with the doc, but I don't
    understand.

    -- 
    Priv.-Doz. Dr. David Meyer
    Institut für Wirtschaftsinformatik

    Fachhochschule Technikum Wien
    Höchstädtplatz 5, 1200 Wien
    T: +43 1 333 40 77-394 <tel:%2B43%201%20333%2040%2077-394>
    F: +43 1 333 40 77-99 394 <tel:%2B43%201%20333%2040%2077-99%20394>
    E: david.me...@technikum-wien.at
    <mailto:david.me...@technikum-wien.at>
    I: www.technikum-wien.at <http://www.technikum-wien.at>



--
Priv.-Doz. Dr. David Meyer
Institut für Wirtschaftsinformatik

Fachhochschule Technikum Wien
Höchstädtplatz 5, 1200 Wien
T: +43 1 333 40 77-394
F: +43 1 333 40 77-99 394
E: david.me...@technikum-wien.at
I: www.technikum-wien.at

______________________________________________
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.