I think that you need to instatiate one of these
https://github.com/usnistgov/fipy/blob/develop/fipy/meshes/mesh2D.py#L70
with the correct arrays. For examples,
~~~~
>>> points = [[0,0], [0, 1], [1, 0], [1, 1]]
>>> from scipy.spatial import Delaunay
>>> tri = Delaunay(points)
>>> tri.points
array([[ 0., 0.],
[ 0., 1.],
[ 1., 0.],
[ 1., 1.]])
>>> tri.simplices
array([[3, 1, 0],
[2, 3, 0]], dtype=int32)
~~~~
So the vertices are just the points but reshaped to (2, N) from (N,
2). The face to vertex array would be
[[3, 1, 0, 2, 0
[1, 0, 3, 3, 2]]
Its all the faces or edges in 2D, and the cell to face ID would be
[[0, 3],
[1, 2],
[2, 4]]
Then you can instatiate the Mesh2D object with those three arrrays. Look at
https://github.com/usnistgov/fipy/blob/develop/fipy/meshes/mesh2D.py#L269
to see how to do it as well. Notice that the test example has both
quads and triangles (the -1 values are masked).
On Wed, Jun 15, 2016 at 1:29 PM, James Pringle <jprin...@unh.edu> wrote:
> Well, I am motivated to give it a go, since I only have the summer to make
> progress on project and it is blocking my research progress right now. Can
> you give me a pointer to where the appropriate quantities are defined? I can
> certainly write code to make the transformations, but it is a bit hard
> without understanding precisely what is defined in the Mesh2D object. I have
> made a simple Mesh2D object, but I am not sure which of the attributes, etc,
> are absolutely required, or how they are precisely defined.
>
> Perhaps most useful for me would be
>
> a definition of which parts of the Mesh2D object must exist
> and the format of those parts, in particular the a face to vertex array and
> a
> cell to face array
>
> Or you could point me to the appropriate part of the Gmsh code so I can go
> from there. I presume poking around in fipy.Gmsh2D would be a good place to
> start? Is there a better place to start?
>
> I would love any documentation on the details of the Mesh2D object.
>
> Thanks,
> Jamie Pringle
> University of New Hampshire
>
> On Wed, Jun 15, 2016 at 12:21 PM, Daniel Wheeler <daniel.wheel...@gmail.com>
> wrote:
>>
>> Hi Jamie,
>>
>> There is no automated way to make a FiPy mesh from Scipy's Delaunay
>> object. The problem is that FiPy needs a face to vertex array and a
>> cell to face array while the Delaunay object just has the cell to
>> vertex array. The functionality to extract the face to vertex array
>> and cell to face array is in FiPy because it must be done for Gmsh,
>> however, it is not abstracted in so that it can be reused.
>>
>> It is certainly possible to make the correct arrays with Numpy and
>> pass them to the Mesh2D object, but it's a bit of work to write the
>> code. If I find some time I might give it a go, but I don't know when
>> I will get to it.
>>
>> Cheers,
>>
>> Daniel
>>
>> On Tue, Jun 14, 2016 at 4:15 PM, James Pringle <jprin...@unh.edu> wrote:
>> > Daniel et al. --
>> >
>> > As referenced earlier in this thread, I have a complex domain with
>> > holes
>> > in it; I now have it broken up into triangles, based on the Delaunay
>> > package
>> > in SciPy. I have
>> >
>> > Locations of all vertex coordinates,
>> > list of vertices that make up faces
>> > list of faces that make cells
>> > list of faces that make up all the internal and external boundaries.
>> >
>> > Can you point me to any code or documentation that would help me
>> > understand
>> > how to make this into a Mesh2D object? I am having a devil of a time
>> > figuring out from the manual online. The best would be something that
>> > used
>> > the output of either the triangles or scipy.spatial.Delaunay() packaged.
>> > My
>> > equation is of the form
>> >
>> > 0=J(Psi,A(x,y)) + \Del(B(x,y)*\Del Psi)
>> >
>> >
>> > and I can get the coefficients A(x,y) and B(x,y) on either the faces or
>> > in
>> > the cell centers are needed.
>> >
>> > Thank you.
>> > Jamie Pringle
>> > University of New Hampshire
>>
>> --
>> Daniel Wheeler
>> _______________________________________________
>> fipy mailing list
>> fipy@nist.gov
>>
>> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.ctcms.nist.gov_fipy&d=CwICAg&c=c6MrceVCY5m5A_KAUkrdoA&r=oMf1-WHpUHeD7kN3S7612CDdF2TDPqDF9R-n-71Ks1Y&m=Q6M6XDjzGDlxMDgrY-riZ7Q70Abl5Aq7SRrTwZNYsws&s=yiO9QAq_3EUy9lVY5F76Tc9mnaB5Ubclilpum_QCUOE&e=
>> [ NIST internal ONLY:
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__email.nist.gov_mailman_listinfo_fipy&d=CwICAg&c=c6MrceVCY5m5A_KAUkrdoA&r=oMf1-WHpUHeD7kN3S7612CDdF2TDPqDF9R-n-71Ks1Y&m=Q6M6XDjzGDlxMDgrY-riZ7Q70Abl5Aq7SRrTwZNYsws&s=yIJJw-iQknimZfYCRrZmk7V2eHSJrCtcAehgVbGeiDk&e=
>> ]
>
>
>
> _______________________________________________
> fipy mailing list
> fipy@nist.gov
> http://www.ctcms.nist.gov/fipy
> [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
>
--
Daniel Wheeler
_______________________________________________
fipy mailing list
fipy@nist.gov
http://www.ctcms.nist.gov/fipy
[ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
