From: <HIDDEN>
Newsgroups: comp.soft-sys.matlab
Subject: Re: Sparse matrix indexing speed -- shape sensitivity
Date: Mon, 29 Aug 2011 10:30:07 -0400
Organization: MathWorks
Lines: 41
Message-ID: <j3g7pf$m8m$>
References: <j3bumn$9ts$> <j3cqer$o66$> <j3df6i$jnq$> <j3dftj$lkh$> <j3dk7r$3mp$> <j3dtcs$rjq$> <j3efp3$h5h$>
Mime-Version: 1.0
Content-Type: text/plain;
Content-Transfer-Encoding: 8bit
X-Trace: 1314628208 22806 (29 Aug 2011 14:30:08 GMT)
NNTP-Posting-Date: Mon, 29 Aug 2011 14:30:08 +0000 (UTC)
In-Reply-To: <HIDDEN>
X-Priority: 3
X-MSMail-Priority: Normal
Importance: Normal
X-Newsreader: Microsoft Windows Live Mail 14.0.8089.726
X-MimeOLE: Produced By Microsoft MimeOLE V14.0.8089.726
Xref: comp.soft-sys.matlab:741588

"Matt J " <mattjacREMOVE@THISieee.spam> wrote in message 
> "Bruno Luong" <b.luong@fogale.findmycountry> wrote in message 
> <j3dtcs$rjq$>...
>> 2) As I understand they way ndSparse combine the n-1 index to 1 linear 
>> index is a limitation. Are you able to store an ndSparse of dimension 
>> [1e6 1e6 1e6 10] (even an empty one)?
> =================
> No, it's true, I cannot. But of course, MATLAB already has limitations 
> similar to this presently on its 2D sparse matrices which user's must work 
> around. Why should it be that this all-zero matrix stores perfectly fine
> A=sparse(1e8,1);
> but not its transpose,
>>> A.';
> ??? Error using ==> transpose
> Out of memory. Type HELP MEMORY for your options.
> Anyway, if extending the indexing supported by setsparse/getsparse is not 
> easy to do, it's not worth the effort.

In the User Guide discussing sparse, there is this statement:

For more information on how MATLAB stores sparse matrices, see John R. 
Gilbert, Cleve Moler , and Robert Schreiber's Sparse Matrices In Matlab: 
Design and Implementation, (SIAM Journal on Matrix Analysis and Applications 
, 13:1, 333356 (1992)).  (link: