dolfin team mailing list archive

[Harish Narayanan <harish.mlists@xxxxxxxxx>]

Anders Logg wrote:
> On Thu, Oct 08, 2009 at 06:46:12PM +0100, Garth N. Wells wrote:
>>
>> Harish Narayanan wrote:
>>> Anders Logg wrote:
>>>> On Thu, Oct 08, 2009 at 05:47:09PM +0200, Anders Logg wrote:
>>>>> On Thu, Oct 08, 2009 at 04:37:14PM +0100, Garth N. Wells wrote:
>>>>>> Anders Logg wrote:
>>>>>>> Failing demos:
>>>>>>>
>>>>>>>   demo/la/eigensolver/cpp (C++)
>>>>>>>
>>>>>>>   demo/pde/dg/advection-diffusion/cpp (C++)
>>>>>>>   demo/pde/dg/advection-diffusion/python (Python)
>>>>>>>
>>>>>>>   demo/pde/dg/poisson/cpp (C++)
>>>>>>>   demo/pde/dg/poisson/python (Python)
>>>>>>>
>>>>>>>   demo/pde/dg/biharmonic/python (Python)
>>>>>>>
>>>>>> I thought I just fixed the DG demos?
>>>>> Yes, works for me now. You fixed it while I ran the tests. :-)
>>>>>
>>>>> And the eigensolver demo seems to work now. Did you fix this?
>>>>>
>>>>> Anyway, I'll rerun the tests now and see if it goes through.
>>>> Everything works here now.
>>> I just wanted to point out that the hyperelasticity demo does not
>>> converge quadratically anymore and prints the following warning during
>>> the Newton-Raphson solve:
>>>
>>> Solving nonlinear variational problem
>>>   Starting Newton solve.
>>>     *** Warning: Invalid value rank of Function 1, got 0 but expecting
>>> 1. You may need to provide the rank of a user defined Function.
>>>     *** Warning: Invalid value rank of Function 2, got 0 but expecting
>>> 1. You may need to provide the rank of a user defined Function.
>>>     Applying boundary conditions to linear system.
>>>     Applying boundary conditions to linear system.
>>>     Matrix of size 2187 x 2187 has 81873 nonzero entries.
>>>     Applying boundary conditions to linear system.
>>>     Applying boundary conditions to linear system.
>>>     Solving linear system of size 2187 x 2187 (PETSc LU solver, umfpack).
>>>
>>> The .h files seems to have value_rank() defined.
>>>
>> The user Expressions probably need to be updated with the rank data.
> 
> Indeed. Should be fixed now.

Thanks!

Harish