Re: [Libmesh-users] errors in ex19 with enabling complex

[Derek Gaston <friedmud@gm...>]

Hmmm... I suppose that's my fault.... sorry about that.  Yujie, does  
just changing Real to Number fix it?

Derek

On Sep 29, 2009, at 3:43 PM, John Peterson wrote:

> On Tue, Sep 29, 2009 at 4:39 PM, Yujie <recrusader@...> wrote:
>> Dear Libmesh Developers,
>>
>> I downloaded the latest Libmesh using SVN. I can compile it with  
>> complex
>> numbers. when I ran "Make run_examples", I got the following errors  
>> for
>> ex19:
>
> It looks like ex19 explicitly uses type Real, when it should probably
> use Number instead (see ex13).
>
> -- 
> John
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry&reg; Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart  
> your
> developing skills, take BlackBerry mobile applications to market and  
> stay
> ahead of the curve. Join us from November 9&#45;12, 2009. Register  
> now&#33;
> http://p.sf.net/sfu/devconf
> _______________________________________________
> Libmesh-users mailing list
> Libmesh-users@...
> https://lists.sourceforge.net/lists/listinfo/libmesh-users

[Libmesh-users] errors in ex19 with enabling complex

[Yujie <recrusader@gm...>]

Dear Libmesh Developers,

I downloaded the latest Libmesh using SVN. I can compile it with complex
numbers. when I ran "Make run_examples", I got the following errors for
ex19:

"make[2]: Leaving directory
`/home/yujie/codes/libmesh_svn_092909/examples/ex18'
make[2]: Entering directory
`/home/yujie/codes/libmesh_svn_092909/examples/ex19'
Compiling C++ (in debug mode) ex19.C...
ex19.C: In function ‘void compute_residual(const
NumericVector<std::complex<double> >&, NumericVector<std::complex<double>
>&)’:
ex19.C:332: error: no match for ‘operator+=’ in ‘u += std::operator* [with
_Tp = double](((const double&)((const double*)((const
std::__debug::vector<double, std::allocator<double> >*)((const
std::__debug::vector<std::__debug::vector<double, std::allocator<double> >,
std::allocator<std::__debug::vector<double, std::allocator<double> > >
>*)phi)->std::__debug::vector<_Tp, _Allocator>::operator[] [with _Tp =
std::__debug::vector<double, std::allocator<double> >, _Allocator =
std::allocator<std::__debug::vector<double, std::allocator<double> >
>](((long unsigned int)j)))->std::__debug::vector<_Tp,
_Allocator>::operator[] [with _Tp = double, _Allocator =
std::allocator<double>](((long unsigned int)qp)))), ((const
std::complex<double>&)((const std::complex<double>*)(&((const
NumericVector<std::complex<double> >*)soln)->NumericVector<T>::operator()
[with T = std::complex<double>](dof_indices. std::__debug::vector<_Tp,
_Allocator>::operator[] [with _Tp = unsigned int, _Allocator =
std::allocator<unsigned int>](((long unsigned int)j)))))))’
/home/yujie/codes/libmesh_svn_092909/include/numerics/type_vector.h: In
member function ‘void TypeVector<T>::add(const TypeVector<T2>&) [with T2 =
std::complex<double>, T = double]’:
/home/yujie/codes/libmesh_svn_092909/include/numerics/type_vector.h:416:
instantiated from ‘const TypeVector<T>& TypeVector<T>::operator+=(const
TypeVector<T2>&) [with T2 = std::complex<double>, T = double]’
ex19.C:188:   instantiated from here
/home/yujie/codes/libmesh_svn_092909/include/numerics/type_vector.h:438:
error: no match for ‘operator+=’ in
‘((TypeVector<double>*)this)->TypeVector<double>::_coords[0] +=
p->TypeVector<std::complex<double> >::_coords[0]’
/home/yujie/codes/libmesh_svn_092909/include/numerics/type_vector.h:439:
error: no match for ‘operator+=’ in
‘((TypeVector<double>*)this)->TypeVector<double>::_coords[1] +=
p->TypeVector<std::complex<double> >::_coords[1]’
/home/yujie/codes/libmesh_svn_092909/include/numerics/type_vector.h:440:
error: no match for ‘operator+=’ in
‘((TypeVector<double>*)this)->TypeVector<double>::_coords[2] +=
p->TypeVector<std::complex<double> >::_coords[2]’
make[2]: *** [ex19.x86_64-unknown-linux-gnu.dbg.o] Error 1
make[2]: Leaving directory
`/home/yujie/codes/libmesh_svn_092909/examples/ex19'
make[1]: *** [run] Error 1
make[1]: Leaving directory `/home/yujie/codes/libmesh_svn_092909/examples'
make: *** [run_examples] Error 2 "

Could you give me some help? Thanks a lot.

Regards,
Yujie

Re: [Libmesh-users] errors in ex19 with enabling complex

[John Peterson <peterson@cf...>]

On Tue, Sep 29, 2009 at 4:39 PM, Yujie <recrusader@...> wrote:
> Dear Libmesh Developers,
>
> I downloaded the latest Libmesh using SVN. I can compile it with complex
> numbers. when I ran "Make run_examples", I got the following errors for
> ex19:

It looks like ex19 explicitly uses type Real, when it should probably
use Number instead (see ex13).

-- 
John

Re: [Libmesh-users] errors in ex19 with enabling complex

[Derek Gaston <friedmud@gm...>]

Hmmm... I suppose that's my fault.... sorry about that.  Yujie, does  
just changing Real to Number fix it?

Derek

On Sep 29, 2009, at 3:43 PM, John Peterson wrote:

> On Tue, Sep 29, 2009 at 4:39 PM, Yujie <recrusader@...> wrote:
>> Dear Libmesh Developers,
>>
>> I downloaded the latest Libmesh using SVN. I can compile it with  
>> complex
>> numbers. when I ran "Make run_examples", I got the following errors  
>> for
>> ex19:
>
> It looks like ex19 explicitly uses type Real, when it should probably
> use Number instead (see ex13).
>
> -- 
> John
>
> ------------------------------------------------------------------------------
> Come build with us! The BlackBerry&reg; Developer Conference in SF, CA
> is the only developer event you need to attend this year. Jumpstart  
> your
> developing skills, take BlackBerry mobile applications to market and  
> stay
> ahead of the curve. Join us from November 9&#45;12, 2009. Register  
> now&#33;
> http://p.sf.net/sfu/devconf
> _______________________________________________
> Libmesh-users mailing list
> Libmesh-users@...
> https://lists.sourceforge.net/lists/listinfo/libmesh-users

Re: [Libmesh-users] errors in ex19 with enabling complex

[Yujie <recrusader@gm...>]

Dear John and Derek,

Thank you for your help.
I adjust "RealGradient" and "Real" in Lines 185, 190, 327, 328, 336 to
"Gradient" and "Number" respectively. I can make it successfully. However, I
don't know whether it is correct. I have not used Libmesh for a long time.

Regards,
Yujie


On Tue, Sep 29, 2009 at 4:47 PM, Derek Gaston <friedmud@...> wrote:

> Hmmm... I suppose that's my fault.... sorry about that.  Yujie, does just
> changing Real to Number fix it?
>
> Derek
>
>
> On Sep 29, 2009, at 3:43 PM, John Peterson wrote:
>
>  On Tue, Sep 29, 2009 at 4:39 PM, Yujie <recrusader@...> wrote:
>>
>>> Dear Libmesh Developers,
>>>
>>> I downloaded the latest Libmesh using SVN. I can compile it with complex
>>> numbers. when I ran "Make run_examples", I got the following errors for
>>> ex19:
>>>
>>
>> It looks like ex19 explicitly uses type Real, when it should probably
>> use Number instead (see ex13).
>>
>> --
>> John
>>
>>
>> ------------------------------------------------------------------------------
>> Come build with us! The BlackBerry&reg; Developer Conference in SF, CA
>> is the only developer event you need to attend this year. Jumpstart your
>> developing skills, take BlackBerry mobile applications to market and stay
>> ahead of the curve. Join us from November 9&#45;12, 2009. Register
>> now&#33;
>> http://p.sf.net/sfu/devconf
>> _______________________________________________
>> Libmesh-users mailing list
>> Libmesh-users@...
>> https://lists.sourceforge.net/lists/listinfo/libmesh-users
>>
>
>

Re: [Libmesh-users] errors in ex19 with enabling complex

[John Peterson <peterson@cf...>]

On Tue, Sep 29, 2009 at 5:05 PM, Yujie <recrusader@...> wrote:
> Dear John and Derek,
>
> Thank you for your help.
> I adjust "RealGradient" and "Real" in Lines 185, 190, 327, 328, 336 to
> "Gradient" and "Number" respectively. I can make it successfully. However, I
> don't know whether it is correct. I have not used Libmesh for a long time.

The output should basically be identical to the output produced by a
non-complex build of Libmesh...if you have time and want to try that,
let us know and we can check that change into the library.

-- 
John

Re: [Libmesh-users] errors in ex19 with enabling complex

[Roy Stogner <roystgnr@ic...>]

On Tue, 29 Sep 2009, Yujie wrote:

> I adjust "RealGradient" and "Real" in Lines 185, 190, 327, 328, 336 to
> "Gradient" and "Number" respectively. I can make it successfully. However, I
> don't know whether it is correct. I have not used Libmesh for a long time.

That should be correct, insofar as you've changed everything necessary
to make the example compile but haven't changed anything
unnecessarily.  I'll commit this to svn if Derek hasn't already.

On the other hand, I'm not sure the question "What is the correct
complex-valued formulation of the Laplace-Young equations" is even
meaningful...  I wonder if we can change those boundary conditions and
end up getting a frequency+magnitude solution for
surface-tension-controlled waves, or something equally interesting?
---
Roy