Re: [Getfem-users] slicer_half_space problems (bug or feature?)

[Roman Putanowicz]

> Dear All,
> 
> I have a problem with slicer_half_space : 
> I would like produce a line slice over a 2D mesh.
> Everything works, fields are interpolated, etc.
> but the actual number of edges and nodes is
> twice as it should be. The edges are just doubled
> and have the same orientation.

Hmm, well, the problem was with my way of interpreting the doc 
not with the slicer :-).

In the examples which I investigated the halfspace boundary
was coincident with the mesh edges thus obviously respective
portions of the boundary were covered by edges of two
triangles. The problem was that I somehow got the impression
that the orientation of the edges should come from the
sliced mesh which is of course wrong -- it comes from 
he orientation of the halfspace boundary.

The second problem was with with the method :
  void  stored_mesh_slice::merge_nodes()

The documentation says that it:
 "build a list of merged nodes" 

It does as promised,  but this doesn't mean that the stored slice mesh
is modified. Probably interpreted the doc in the wrong way by looking at the 
return value which is void. Here come at last the positive outcome
of the whole story: 
  ** there is a set of functions in stored_mesh_slice which provide
interface to merged nodes. Simply, they are not documented (which I will
try to fix :). This explain also why mrege_nodes is of void type.

I am sorry for this false alarm. 

Best regards,

Roman
-- 
Roman Putanowicz, PhD  < address@hidden  >
Institute for Computational Civil Engng (L-5)
Dept. of Civil Engng, Cracow Univ. of Technology
www.l5.pk.edu.pl, tel. +48 12 628 2569, fax 2034