Re: [eigen] New Levenberg Marquardt stuff

[ Thread Index | Date Index | More lists.tuxfamily.org/eigen Archives ]

Hi,

I totally agree with Sameer here. Clearly Ceres is too sophisticated to be part of Eigen. Moreover, it's for double only, while we strive to stay as generic as possible.

gael 


On Fri, Dec 7, 2012 at 11:18 PM, Sameer Agarwal <sameeragarwal@xxxxxxxxxx> wrote:
No, Its not a good idea.

Eigen is a linear algebra library (or atleast thats how most people think about it).
why should a full fledged nonlinear least squares library be included in it?
based on eigen is a very  loose term. we use eigen extensively, but thats not the only thing we use.
let me rephrase, Ceres is a large and complicated dependency.

Sameer



On Fri, Dec 7, 2012 at 1:57 PM, Thomas Capricelli <orzel@xxxxxxxxxxxxxxx> wrote:
Il 07/12/2012 22:17, Sameer Agarwal ha scritto:
> Ceres is a rather large library, putting it into Eigen is neither
> practical or a good idea.

If it has a better LM, and if that LM is based on eigen, it would
definitely be a great option to move it to eigen, don't you think ? The
same for the other generic algorithm there might be there.



Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/