(PR#7045) Re: [Rd] homals: Error in y[g[i, j], ] : incorrect number

[ripley at stats.ox.ac.uk]

Function homals() is not part of R.  It appears to be in a contributed 
package `homals', but you did not say so.

To quote the FAQ

   Bug reports on contributed packages should be sent first to the package
   maintainer, and only submitted to the R-bugs repository by package
   maintainers, mentioning the package in the subject line.

I have closed this bug report on R-bugs.


On Fri, 2 Jul 2004 Michael.G.Shafto at nasa.gov wrote:

> Full_Name: Michael G. Shafto
> Version: 1.9.1
> OS: Microsoft Windows XP
> Submission from: (NULL) (192.150.38.75)
> 
> 
> > homals(x, sets=0, ndim=3, active=T, rank=3, level="NO", starplots=FALSE,
> + catplots=FALSE, trfplots=FALSE, lossplots=FALSE, hullplots=FALSE,
> + spanplots=FALSE, graphplot=TRUE, objplot=TRUE, objscores=TRUE,
> + objlabel=TRUE, offset=1.20, eps1 = -Inf, eps2=1e-5, itermax=32,
> + voronoi=FALSE, saveMe=TRUE, demo=FALSE, timer=FALSE, tk=FALSE)
> Iteration:    1  Eigenvalues:   0.275420  0.182711  0.117240  Gain:   0.575371 
> 0.981319 
> Iteration:    2  Eigenvalues:   0.284051  0.185338  0.129728  Gain:   0.599117 
> 0.992531 
> Iteration:    3  Eigenvalues:   0.285723  0.186139  0.137560  Gain:   0.609422 
> 0.995753 
> 
> ... etc. ...
> 
> Iteration:   26  Eigenvalues:   0.286144  0.186919  0.155740  Gain:   0.628804 
> 0.999988 
> Iteration:   27  Eigenvalues:   0.286144  0.186919  0.155761  Gain:   0.628824 
> 0.999990 
> Iteration:   28  Eigenvalues:   0.286144  0.186919  0.155778  Gain:   0.628841 
> 0.999991 
> Error in y[g[i, j], ] : incorrect number of dimensions
> ------------------------------------------------------
> I get this same error with various different parameter settings in homals.
> For example,
>      homals(x, graphplot = TRUE, objplot = TRUE)
> produces the same error message.
> 
> I am using a rather large data set. 2237 rows and 1196 columns. 
> Also, some variables have range=0 (all cases have the same value).
> Still, the error message has no clear relation to either of these properties
> of the data.
> The Graphplot for dimension 1 vs. dimension 2 is produced in a readable .pdf
> file.

-- 
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