Re: blog update #2

[c.]

On 18 Jun 2013, at 00:07, Marco Vassallo <address@hidden> wrote:

> The problem at the moment is that I don't know to which side of the dolfin 
> mesh the i-th column
> of the e matrix corresponds. 
> In fact, I'm adding the information to the e matrix iterating over all the 
> edges and if they are on the boundary I add them to "e".
> But, on the other side, the additional information stored in dolfin xml is 
> something like 
> <value cell_index="21997" local_entity="0" value="0"/>
> but I don't know where I stored the information corresponding to the cell = 
> 21997, for example. 
> The map I have in mind is just something which says the correspondance 
> between the elements of the two meshes.

I really don't understand what you are saying here, all sems very straight 
forward to me, 
so either I completely misinterpreted the meaning of the data in the dolfin xml 
file or you 
are very confused about what we should be trying to do.

Can you right in the blog a post explaining what exactly is the meaning of each 
part of:

   <value cell_index="21997" local_entity="0" value="0"/>

? you should ask more info to the FENICS people if you have doubts.
c.