Re: New function: lsqnonlin

[Ben Abbott]

On Apr 2, 2008, at 2:54 PM, Jonathan Stickel wrote:
> On 4/2/08 address@hidden wrote:
> > Date: Wed, 02 Apr 2008 14:17:14 -0400
> > From: "John W. Eaton" <address@hidden>
> > Subject: New function: lsqnonlin
> > To: Bill Denney <address@hidden>
> > Cc: octave maintainers mailing list <address@hidden>
> > Message-ID: <address@hidden>
> > Content-Type: text/plain; charset=us-ascii
> > On  1-Apr-2008, Bill Denney wrote:
> > | Attached is lsqnonlin.
> > I applied this changeset.  Please remember to add new functions to  
> > the
> > SOURCES list in the correspoding Makefile.in file.
>
> How does this function compare to "leasqr" in octave-forge?
>
> Jonathan

The title to the thread is incorrectly labeled. It should read  
"lsqnonneg", not "lsqnonlin"

In any event, the principle difference is that lsqnonneg constrains  
the solution to non-negative values.

From the "other program's" help ;-)

 LSQNONNEG Linear least squares with nonnegativity constraints.
    X = LSQNONNEG(C,d) returns the vector X that minimizes NORM(d-C*X)
    subject to X >= 0. C and d must be real.

Ben