[R] converting lists got by tapply to dataframes

[Prof Brian Ripley]

On Wed, 10 Mar 2004, Liaw, Andy wrote:

> That gives you a vector, not a data frame.
> 
> The problem is that the object returned by tapply is a 1-d array of list.
> When as.data.frame() is called, the default method doesn't know how to
> handle arrays and gives the error.  What you really want for this particular
> case is to treat the object as a list, so the unorthodox way is:
> 
>   as.data.frame.list(xb)
> 
> OK, I didn't really say that...

If this dataset barley from package lattice, it works in 1.9.0 alpha
which does have as.data.frame.array.  So the issue is about to go away.


> > On Wednesday 10 March 2004 06:58, Wolfram Fischer wrote:
> > > I have two lists:
> > >
> > >     xa <- list( X=c(1,2,3), Y=c(4,5,6), Z=c(7,8,9) )
> > >
> > >     xb <- with( barley, tapply( X=seq(1:nrow(barley)), INDEX=site
> > >         , FUN=function(z)yield[z]))
> > >
> > > I can convert xa to a dataframe easily with:
> > >     as.data.frame(xa)
> > >
> > > But if i try the same with xb I get:
> > >     as.data.frame(xb)
> > >     Error in as.data.frame.default(xb) :
> > >     can't coerce array into a data.frame
> > >
> > > What helps?
> > >
> > > (NB: I know the formula for xb is stupid, but it generates the same
> > > type of list as the list I get from my real problem.)
> 

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595