Dear Rolnd and Zach,

Thanks for your explanation. My domain size is (-540,-540,-540)x(540,540,540), so the interpolation point is outside of the domain as you said.

For using out_of_bounds_xyz, do I need to rerun the simulation from the beginning or I can run it from the last checkpoint after adding out_of_bounds_xyz in the parameter file ?

Best regards,

Chia-Hui

________________________________ 寄件者: Zach Etienne zachetie@gmail.com 寄件日期: 2019年5月30日 下午 11:03:03 收件者: Haas, Roland 副本: 林家暉; Einstein Toolkit Users 主旨: Re: [Users] interpolation error

Hi Chia-Hui,

particle_tracerET already has a parameter for this: out_of_bounds_xyz

I follow the general rule of setting it to about 80% of my (coarsest) grid size.

-Zach

* * * Prof. Zachariah Etienne West Virginia University https://math.wvu.edu/~zetienne/ https://blackholesathome.net

On Thu, May 30, 2019 at 10:58 AM Haas, Roland <rhaas@illinois.edumailto:rhaas@illinois.edu> wrote: Hello Chia-Hui,

the error message indicates that the code tried to interpolation a quantity to location (x,y,z)=(-157.725,-576.349,-99.2027) but that the grid only has a size of (xmin,ymin,zmin)x(xmax,ymax,zmax) = (-576,-576,-576)x(576,576,576) (where I made a guess for the ymax and zmax values given that the output only shows the shape of the grid on one MPI rank).

So the interpolation point is outside of the domain.

Depending on what caused the interpolation this could mean eg that a tracer particle (given the function name "DoOneRK4StepForParticleTracerET") tried to leave the domain. This is something that needsto be fixed in the particle_tracerET thorn, probably by making it deactivate particles that get too close to the edge of the domain.

Yours, Roland

Hello,

I am using IllinoisGRMHD with the test with nsns_test.par in https://bitbucket.org/zach_etienne/wvuthorns_diagnostics/src/master/NSNS_par... .

The simulation runs well until t~5300 and an error occurs. The error message is like this:

WARNING level 1 from host nid02210 process 0 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn AEILocalInterp, file /global/u1/c/chlin/et/Cactus/arrangements/Numerical/AEILocalInterp/src/Lagrange-tensor-product/../template.c:1109: -> CCTK_InterpLocalUniform(): interpolation point is either outside the grid, or inside but too close to the grid boundary! (this may be caused by a global interpolation with driver::ghost_size too small) 0-origin interpolation point number pt=0 of N_interp_points=1 interpolation point (x,y,z)=(-157.725,-576.349,-99.2027) grid x_min(delta_x)x_max = -576(18)576 grid y_min(delta_y)y_max = -576(18)54 grid z_min(delta_z)z_max = -576(18)54

WARNING level 1 from host nid02210 process 0 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn particle_tracerET, file /global/u1/c/chlin/et/Cactus/arrangements/WVUThorns_Diagnostics/particle_tracerET/src/Interpolate_velocities_at_particle_positions.C:131: -> interpolation screwed up WARNING level 1 from host nid02210 process 1 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn particle_tracerET, file /global/u1/c/chlin/et/Cactus/arrangements/WVUThorns_Diagnostics/particle_tracerET/src/Interpolate_velocities_at_particle_positions.C:131: -> interpolation screwed up srun: error: nid02211: tasks 2-3: Exited with exit code 1 srun: Terminating job step 21753671.0 srun: error: nid02210: tasks 0-1: Exited with exit code 1

I think maybe t can be avoided by decreasing the timestep so that the tracer particle would not move too fast to make the interpolation screwed up but I am not sure. So what I want to ask is :

1. It this error can be avoided by decreasing the timestep ?
2. If so, how can I decrease the timestep in parameter file (or in other places)?

Thank you. Best regards, Chia-Hui

-- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://pgp.mit.edu . _______________________________________________ Users mailing list Users@einsteintoolkit.orgmailto:Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users

Dear Zach,

I see. Thank you !

Best regards,

Chia-Hui

________________________________ 寄件者: Zach Etienne zachetie@gmail.com 寄件日期: 2019年5月31日 上午 12:47:00 收件者: 林家暉 副本: Haas, Roland; Einstein Toolkit Users 主旨: Re: [Users] interpolation error

Dear Chia-Hui,

out_of_bounds_xyz is denoted as STEERABLE=ALWAYS. Therefore you can change the value at any time in your parameter file, even after a checkpoint.

-Zach

* * * Prof. Zachariah Etienne West Virginia University https://math.wvu.edu/~zetienne/ https://blackholesathome.net

On Thu, May 30, 2019 at 11:43 AM 林家暉 <r06222015@ntu.edu.twmailto:r06222015@ntu.edu.tw> wrote:

Dear Rolnd and Zach,

Thanks for your explanation. My domain size is (-540,-540,-540)x(540,540,540), so the interpolation point is outside of the domain as you said.

For using out_of_bounds_xyz, do I need to rerun the simulation from the beginning or I can run it from the last checkpoint after adding out_of_bounds_xyz in the parameter file ?

Best regards,

Chia-Hui

________________________________ 寄件者: Zach Etienne <zachetie@gmail.commailto:zachetie@gmail.com> 寄件日期: 2019年5月30日 下午 11:03:03 收件者: Haas, Roland 副本: 林家暉; Einstein Toolkit Users 主旨: Re: [Users] interpolation error

Hi Chia-Hui,

particle_tracerET already has a parameter for this: out_of_bounds_xyz

I follow the general rule of setting it to about 80% of my (coarsest) grid size.

-Zach

* * * Prof. Zachariah Etienne West Virginia University https://math.wvu.edu/~zetienne/ https://blackholesathome.net

On Thu, May 30, 2019 at 10:58 AM Haas, Roland <rhaas@illinois.edumailto:rhaas@illinois.edu> wrote: Hello Chia-Hui,

the error message indicates that the code tried to interpolation a quantity to location (x,y,z)=(-157.725,-576.349,-99.2027) but that the grid only has a size of (xmin,ymin,zmin)x(xmax,ymax,zmax) = (-576,-576,-576)x(576,576,576) (where I made a guess for the ymax and zmax values given that the output only shows the shape of the grid on one MPI rank).

So the interpolation point is outside of the domain.

Depending on what caused the interpolation this could mean eg that a tracer particle (given the function name "DoOneRK4StepForParticleTracerET") tried to leave the domain. This is something that needsto be fixed in the particle_tracerET thorn, probably by making it deactivate particles that get too close to the edge of the domain.

Yours, Roland

Hello,

I am using IllinoisGRMHD with the test with nsns_test.par in https://bitbucket.org/zach_etienne/wvuthorns_diagnostics/src/master/NSNS_par... .

The simulation runs well until t~5300 and an error occurs. The error message is like this:

WARNING level 1 from host nid02210 process 0 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn AEILocalInterp, file /global/u1/c/chlin/et/Cactus/arrangements/Numerical/AEILocalInterp/src/Lagrange-tensor-product/../template.c:1109: -> CCTK_InterpLocalUniform(): interpolation point is either outside the grid, or inside but too close to the grid boundary! (this may be caused by a global interpolation with driver::ghost_size too small) 0-origin interpolation point number pt=0 of N_interp_points=1 interpolation point (x,y,z)=(-157.725,-576.349,-99.2027) grid x_min(delta_x)x_max = -576(18)576 grid y_min(delta_y)y_max = -576(18)54 grid z_min(delta_z)z_max = -576(18)54

WARNING level 1 from host nid02210 process 0 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn particle_tracerET, file /global/u1/c/chlin/et/Cactus/arrangements/WVUThorns_Diagnostics/particle_tracerET/src/Interpolate_velocities_at_particle_positions.C:131: -> interpolation screwed up WARNING level 1 from host nid02210 process 1 while executing schedule bin particle_tracerET, routine particle_tracerET::DoOneRK4StepForParticleTracerET in thorn particle_tracerET, file /global/u1/c/chlin/et/Cactus/arrangements/WVUThorns_Diagnostics/particle_tracerET/src/Interpolate_velocities_at_particle_positions.C:131: -> interpolation screwed up srun: error: nid02211: tasks 2-3: Exited with exit code 1 srun: Terminating job step 21753671.0 srun: error: nid02210: tasks 0-1: Exited with exit code 1

I think maybe t can be avoided by decreasing the timestep so that the tracer particle would not move too fast to make the interpolation screwed up but I am not sure. So what I want to ask is :

1. It this error can be avoided by decreasing the timestep ?
2. If so, how can I decrease the timestep in parameter file (or in other places)?

Thank you. Best regards, Chia-Hui

-- My email is as private as my paper mail. I therefore support encrypting and signing email messages. Get my PGP key from http://pgp.mit.edu . _______________________________________________ Users mailing list Users@einsteintoolkit.orgmailto:Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users