[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ff3d-users] Re: Re: Re: Reading GMSH hexahedral mesh.
From: |
Stephane Del Pino |
Subject: |
Re: [ff3d-users] Re: Re: Re: Reading GMSH hexahedral mesh. |
Date: |
Wed, 14 Mar 2007 07:26:45 +0100 |
User-agent: |
KMail/1.9.5 |
Hello.
Le mardi 13 mars 2007 11:29, Cécile Giorla a écrit :
> FunctionExpressionConvection.cpp:74: convection is not implemented for
> this kind of mesh
> UNEXPECTED ERROR: this should not occure, please report it
Thanks for the report. This requires some work. I will let you know when it is
fixed.
> By the way, I have another question :
> Would you tell me how to store a matrix when solving an elliptic PDE and
> re-use it in another iteration? I've seen there is a "memory" option
> which can be set to "matrix = sparse". But is it possible to extract
> this matrix in order to re-use it (like in Freefem++)?
The option you suggest is used if you want to save memory: the matrix is not
assembled (matrix=none).
There is no way to keep matrix into memory. Again, this optimization is to be
done but require a big amount of work. There are other optimizations which
will be more important to do before...
> Another remark : I have got the cvs version taken yesterday, and the
> GMSH reader is different : it now takes into account the number of the
> elementary entity, and no more the number of physical entity. Physical
> entity is more comprehensive for people, is it possible to go back to
> the labeling of elements through the physical entities?
Thanks for the remark, this is fixed, but I cannot commit it in cvs, since
savannah seems to be down.
Best regards,
Stéphane.