[R] Transform pairwise observations into a table

William Dunlap wdunlap at tibco.com
Tue Oct 2 17:13:21 CEST 2012 

Another base-R-only solution uses a 2-column matrix of subscripts
to fill a matrix.  E.g.,

> f <- function(data) {
+     mat <- matrix(NA_real_, nrow=max(data[[1]]), ncol=max(data[[2]]))
+     mat[cbind(data[[1]], data[[2]])] <- data[[3]]
+     mat
+ }
> 
> f(dat1)
      [,1]  [,2]  [,3] [,4]
[1,] 1.000 0.250 0.125  0.5
[2,] 0.250 1.000 0.125  0.5
[3,] 0.125 0.125 1.000  0.5
[4,] 0.500 0.500 0.500  1.0

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf
> Of David Winsemius
> Sent: Monday, October 01, 2012 7:33 PM
> To: arun
> Cc: R help; Bert Gunter; Hadjixenofontos, Athena
> Subject: Re: [R] Transform pairwise observations into a table
> 
> 
> On Oct 1, 2012, at 2:30 PM, arun wrote:
> 
> > HI AHJ,
> > No problem.
> >
> > One more way in addition to reshape() (Rui's suggestion) to get the same result.
> > library(reshape)
> >
> > as.matrix(cast(melt(dat1,id=c("ind1","ind2")),ind1~ind2,value="value"))
> > #      1     2     3   4
> > #1 1.000 0.250 0.125 0.5
> > #2 0.250 1.000 0.125 0.5
> > #3 0.125 0.125 1.000 0.5
> > #4 0.500 0.500 0.500 1.0
> > A.K.
> >
> That looks a tad ... well, ... complicated. So perhaps these base-only solutions with tapply
> might be more accessible: Some of them do border on the whimsical, I will admit:
> 
> with (dat1, tapply(coef, list(ind1,ind2), I))
> 
> with (dat1, tapply(coef, list(ind1,ind2), c))
> 
> with (dat1, tapply(coef, list(ind1,ind2), "^", 1))
> 
> with (dat1, tapply(coef, list(ind1,ind2), "+", 0))
> 
> It is a specific response to the request for a `table`-like function tha twouldallow the
> application of other functions. Cnage the `1` to `2` in the third instance and you get the
> tabulated squares. And do not forget the availability of `ftable` to flatten the output of
> `tapply` retunred values.
> 
> 
> --
> David.
> >
> >
> > ----- Original Message -----
> > From: "Hadjixenofontos, Athena" <AHadjixenofontos at med.miami.edu>
> > To: arun <smartpink111 at yahoo.com>
> > Cc: R help <r-help at r-project.org>
> > Sent: Monday, October 1, 2012 12:59 PM
> > Subject: Re: [R] Transform pairwise observations into a table
> >
> > Thank you. I had looked at xtabs but misunderstood the syntax. This is great. :)
> >
> > AHJ
> >
> >
> >
> > On Oct 1, 2012, at 12:53 PM, "arun" <smartpink111 at yahoo.com> wrote:
> >
> >> Hi,
> >> Try this:
> >>
> >> dat1<-read.table(text="
> >> ind1 ind2 coef
> >> 1 1 1
> >> 1 2 0.25
> >> 1 3 0.125
> >> 1 4 0.5
> >> 2 2 1
> >> 2 1 0.25
> >> 2 3 0.125
> >> 2 4 0.5
> >> 3 3 1
> >> 3 1 0.125
> >> 3 2 0.125
> >> 3 4 0.5
> >> 4 4 1
> >> 4 1 0.5
> >> 4 2 0.5
> >> 4 3 0.5
> >> ",sep="",header=TRUE)
> >> mat1<-as.matrix(xtabs(coef~ind1+ind2,data=dat1))
> >>
> >> #    ind2
> >> #ind1     1     2     3     4
> >>    # 1 1.000 0.250 0.125 0.500
> >>     #2 0.250 1.000 0.125 0.500
> >>     #3 0.125 0.125 1.000 0.500
> >>     #4 0.500 0.500 0.500 1.000
> >>
> >> A.K.
> >>
> >>
> >>
> >> ----- Original Message -----
> >> From: AHJ <ahadjixenofontos at med.miami.edu>
> >> To: r-help at r-project.org
> >> Cc:
> >> Sent: Monday, October 1, 2012 12:17 PM
> >> Subject: [R] Transform pairwise observations into a table
> >>
> >> Hi,
> >>
> >> I have a table of pairs of individuals and a coefficient that belongs to the
> >> pair:
> >>
> >> ind1    ind2    coef
> >> 1    1    1
> >> 1    2    0.25
> >> 1    3    0.125
> >> 1    4    0.5
> >> 2    2    1
> >> 2    1    0.25
> >> 2    3    0.125
> >> 2    4    0.5
> >> 3    3    1
> >> 3    1    0.125
> >> 3    2    0.125
> >> 3    4    0.5
> >> 4    4    1
> >> 4    1    0.5
> >> 4    2    0.5
> >> 4    3    0.5
> >>
> >> And I want to convert it to a matrix where each individual is both a row and
> >> a column and at the intersection of each pair is the coefficient that
> >> belongs to that pair:
> >>
> >>      1           2           3             4
> >> 1    1           0.25           0.125    0.5
> >> 2    0.25      1           0.125    0.5
> >> 3    0.125    0.125    1            0.5
> >> 4    0.5           0.5           0.5            1
> >>
> >> If table() would allow me to specify something other than frequencies to
> >> fill the table with, it would be what I need. I tried a few different
> >> combinations of t() and unique() but none of it made enough sense to post as
> >> my starting code... I am just lost. Any help would be greatly appreciated.
> >>
> >> Thank you,
> >> AHJ
> >>
> >>
> >>
> >> --
> >> View this message in context: http://r.789695.n4.nabble.com/Transform-pairwise-
> observations-into-a-table-tp4644706.html
> >> Sent from the R help mailing list archive at Nabble.com.
> >>
> >> ______________________________________________
> >> 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, MD
> Alameda, CA, USA
> 
> ______________________________________________
> 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.

More information about the R-help mailing list