From help-request at octave dot org Wed Apr  5 11:52:23 2006
Subject: 'cat'-ing single matrices
From: Joshua Rigler <jrigler at hao dot ucar dot edu>
To: help at octave dot org
Date: Wed, 5 Apr 2006 11:40:46 -0500

Can someone say whether Octave versions greater than 2.1.71 have a 'cat' 
function that does *not* return an error if passed a single matrix?  In 
other words, if one types:

   >> cat(1,rand(3,1))

in 2.1.71 (and earlier, I presume), an error is returned (actually, the 
built-in function currently forces the output of 'help cat', which is a 
little annoying).

It seems to me that this could be a relatively benign (non)feature where 
cat just returns the single variable passed to it originally.  It would 
be helpful in the O-F contributed function cell2mat.m, which is causing 
me problems when I try to convert a cell array of structures to a plain 
old array of structures.  FWIW, I did notice that cell2mat.m has been 
migrated from O-F to Octave since I last upgraded, but it doesn't look 
much different than the version of cell2mat.m I currently have.

It might also be worth noting that I tried slightly modifying cell2mat.m 
as follows:

   # m = cat(1, c1{:}); # changed to...
   m = cat(1, c1{:}, []);

which works fine if c1 contains a single matrix of scalars, but as you 
might already have surmised, when c1 contains a single array of 
structures, Octave-2.1.71 doesn't know what to do with the '[]', and 
gives the following error message:

   error: concatenation operator not implemented for `struct' by 
`matrix' operations

Thanks for any/all advice.

-EJR



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------