It seems that the issue was some bug in the earlier versions of nektar. Everything works fine in the latest master branch version.


Petteri Peltonen, (M.Sc.)
Aalto University, School of Engineering
Department of Mechanical Engineering
PO Box 14300
FI-00076 Aalto, Finland
+358 44 0554984
petteri.peltonen@aalto.fi



Lähettäjä: Douglas Serson <d.serson@gmail.com>
Lähetetty: 24. marraskuuta 2017 16:51
Vastaanottaja: Peltonen Petteri
Kopio: nektar-users@imperial.ac.uk
Aihe: Re: [Nektar-users] interppoints:box in parallel
 
Hi Petteri,

Could you try this on the latest master branch anyway? At least this way we can rule out the possibility that this was a bug from v4.4.0 that was already fixed.

You can find instructions on how to clone the repository on section 1.3 of the user guide. 

Cheers,
Douglas

2017-11-24 6:37 GMT-02:00 Peltonen Petteri <petteri.peltonen@aalto.fi>:

I'm using version 4.4.0 so it's likely something I'm doing.


Petteri Peltonen, (M.Sc.)
Aalto University, School of Engineering
Department of Mechanical Engineering
PO Box 14300
FI-00076 Aalto, Finland
+358 44 0554984
petteri.peltonen@aalto.fi



Lähettäjä: Douglas Serson <d.serson@gmail.com>
Lähetetty: 23. marraskuuta 2017 20:02
Vastaanottaja: Peltonen Petteri
Kopio: nektar-users@imperial.ac.uk
Aihe: Re: [Nektar-users] interppoints:box in parallel
 
Hi Petteri,

What version of Nektar++ are you using?
I tried a similar test on the latest master branch and if worked fine, so it is possible that this is a bug that was recently fixed.

Cheers,
Douglas

2017-11-23 15:15 GMT-02:00 Peltonen Petteri <petteri.peltonen@aalto.fi>:

Hey,


I would like to run a POD on a section of my 3D field on a coarser grid. For this, I would like the section to be equidistributed cartesian box and this is exactly what the '-m interppoints:box' utility does. However, I'm having trouble running the utility in parallel. When I run the command:

FieldConvert -m interppoints:fromxml=base/jet.xml:fromxml=base/mesh.xml:fromfld=base/mesh_1.chk:\
box=10,10,10,-0.001,0.001,-0.001,0.001,-0.003,0.003 box1.dat

the utility does exactly what I want.


When I run it in parallel:

mpirun -np 3 FieldConvert \
-m interppoints:fromxml=base/jet.xml:fromxml=base/mesh.xml:fromfld=base/mesh_1.chk:\
box=10,10,10,-0.001,0.001,-0.001,0.001,-0.003,0.003 box2.dat


box2.dat only contains one third of the points in the box. I also tried adding box2.dat:dat:writemultiplefiles, in which case I get the correct number of points split in three different files, but the file is still somehow corrupted when I open it with paraview. What am I doing wrong?


Petteri Peltonen, (M.Sc.)
Aalto University, School of Engineering
Department of Mechanical Engineering
PO Box 14300
FI-00076 Aalto, Finland
+358 44 0554984
petteri.peltonen@aalto.fi

_______________________________________________
Nektar-users mailing list
Nektar-users@imperial.ac.uk
https://mailman.ic.ac.uk/mailman/listinfo/nektar-users