Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Paul T. Bauman
On Wed, Mar 20, 2013 at 12:56 PM, Paul T. Bauman wrote: > On Wed, Mar 20, 2013 at 12:40 PM, Manav Bhatia wrote: > >> >> Please send me the tips/pointers and I will likely get to it this >> afternoon. >> > > The strategy that I had in mind would be similar to what's done for > vector-valued compon

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Manav Bhatia
Hi Ben, That sounds good. My current application needs a solution with real numbers and then a solution with complex numbers using the real solution as reference. The intent is to write the equation system in the application built with real numbers, and then read it again for the libr

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Paul T. Bauman
Arg, didn't hit reply all. On Wed, Mar 20, 2013 at 12:56 PM, Paul T. Bauman wrote: > On Wed, Mar 20, 2013 at 12:40 PM, Manav Bhatia wrote: > >> >> Please send me the tips/pointers and I will likely get to it this >> afternoon. >> > > The strategy that I had in mind would be similar to what's don

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Kirk, Benjamin (JSC-EG311)
On Mar 20, 2013, at 12:50 PM, Manav Bhatia wrote: > If I am using double precision numbers, then does XDR write binary files > with double precisions? that is correct. > To enable complex IO, the real and imaginary parts could be written as two > separate vectors and then read back as two sepa

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Manav Bhatia
If I am using double precision numbers, then does XDR write binary files with double precisions? To enable complex IO, the real and imaginary parts could be written as two separate vectors and then read back as two separate vectors before combining the values into a complex number? Manav On We

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Manav Bhatia
Yes, I am looking for some solution soon, and can get to it today. Please send me the tips/pointers and I will likely get to it this afternoon. Manav On Wed, Mar 20, 2013 at 1:35 PM, Paul T. Bauman wrote: > On Wed, Mar 20, 2013 at 12:31 PM, Manav Bhatia wrote: > >> Hi, >> >>Which of the

Re: [Libmesh-users] IO for complex numbers

2013-03-20 Thread Paul T. Bauman
On Wed, Mar 20, 2013 at 12:31 PM, Manav Bhatia wrote: > Hi, > >Which of the Input/Output classes (specifically binary files) are > consistent with complex numbers? > Right now, it's only GMV. I was supposed to take care of this a few months ago (for all formats except VTK) and it got put on t

[Libmesh-users] IO for complex numbers

2013-03-20 Thread Manav Bhatia
Hi, Which of the Input/Output classes (specifically binary files) are consistent with complex numbers? I noticed that VTKIO writes only the real part of the number (and it is ascii). I remember seeing some discussion on the mailing list about ExodusII being for real numbers only (please cor