Re: [Libmesh-users] about debugging and parallel

[Roy S. <roy...@ic...>]

On Wed, 5 Sep 2007, Yujie wrote:

> I wrote my code according to example 3(there, quad9 is
> used.). I didn't almost change the code. I use tetrehadron element. I think
> that this change leads to the error. However, I don't know which parameters
> to control the choice you said(FEType fe_type = dof_map.variable_type(0);,
> this one?). Could you give me some advice?

Sorry about the late reply.

The base polynomial degree you use is set by this line in ex3.C:

     equation_systems.get_system("Poisson").add_variable("u", SECOND);

Which tells the system to use second-order approximation.  If you
haven't changed that line, then the error alert in fe_lagrange.C must
have been triggered by a more subtle problem.
---
Roy

>> the original bug here is at "fe_lagrange.C, line 628", and
>> it's basically reporting that you're trying to use LAGRANGE finite
>> elements with something other than linear or quadratic shape
>> functions.  We only support p=1 and p=2 with Lagrange elements; if you
>> need higher you'll have to switch to the Hierarchic elements.
>> ---
>> Roy
>>
>