Trac

trac@lists.einsteintoolkit.org

8708 discussions

#2631: Meudon_BinNS fails if initial_dtshift or initial_dtlapse is set to an unexpected value
by Roland Haas 19 Aug '22

19 Aug '22

#2631: Meudon_BinNS fails if initial_dtshift or initial_dtlapse is set to an unexpected value Reporter: Roland Haas Status: new Milestone: Version: Type: bug Priority: minor Component: Comment (by Roland Haas): One may also have to check if this will not fail when `cctk_lsh != cctk_ash` ie when padding is used. Currently the code uses: ```c++ int const npoints = cctk_lsh[0] * cctk_lsh[1] * cctk_lsh[2]; vector<double> xx(npoints), yy(npoints), zz(npoints); ``` and ``` for (int i=0; i<npoints; ++i) { if (CCTK_EQUALS(initial_lapse, "Meudon_Bin_NS")) { alp[i] = bin_ns.nnn[i]; } } ``` which is incorrect when padding is used. ‌ -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2631/meudon_binns-fail…

1 0

#2631: Meudon_BinNS fails if initial_dtshift or initial_dtlapse is set to an unexpected value
by Roland Haas 19 Aug '22

19 Aug '22

#2631: Meudon_BinNS fails if initial_dtshift or initial_dtlapse is set to an unexpected value Reporter: Roland Haas Status: new Milestone: Version: Type: bug Priority: minor Component: Meudon\_BinNS contains code like this: ‌ ```c if (CCTK_EQUALS(initial_lapse, "Meudon_Bin_NS")) { if (CCTK_EQUALS (initial_dtlapse, "Meudon_Bin_NS")) { CCTK_INFO ("Calculating time derivatives of lapse"); set_dt_from_domega (CCTK_PASS_CTOC, alp, dtalp, omega); } else if (CCTK_EQUALS (initial_dtlapse, "none") or CCTK_EQUALS(initial_dtlapse,"zero")) { // do nothing } else { CCTK_WARN (CCTK_WARN_ABORT, "internal error"); } } ``` which fails if `initial_lapse` is set to `Meudon_Bin_NS` but `initial_dtlapse` is set to an unexpected value \(and similar for `shift`\). It also only sets `dtalp` if it sets `alp`. Neither is correct. The correct way, documented here [https://www.einsteintoolkit.org/thornguide/EinsteinBase/ADMBase/documentati…, is to inspect each parameter independently and then, if the the parameter has a value matching one of the values that `Meudon_Bin_NS` added to the list of allowed keyword values, set the variable accordingly. That is something along the lines: ```c for (int i=0; i<npoints; ++i) { if (CCTK_EQUALS(initial_lapse, "Meudon_Bin_NS")) { alp[i] = bin_ns.nnn[i]; } } if (CCTK_EQUALS (initial_dtlapse, "Meudon_Bin_NS")) { // use bin_ns.nnn to compute derivs of lapse in case alp is not set by us set_dt_from_domega (CCTK_PASS_CTOC, bin_ns.nnn, dtalp, omega); } ``` ‌ -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2631/meudon_binns-fail…

1 0

#2485: include complex and real scalar evolution code from Canuda in ET
by Miguel Zilhão 18 Aug '22

18 Aug '22

#2485: include complex and real scalar evolution code from Canuda in ET Reporter: Roland Haas Status: open Milestone: Version: Type: task Priority: major Component: Comment (by Miguel Zilhão): > ‌ In ScalarEvolve, why BSSN-like variables are public ? Since same name variables are used in other thorn \(like lean\_public\) and they are just auxiliary field to calculate rhs of the equations, private is better for them. fixed in https://bitbucket.org/canuda/scalar/commits/98828a12f314e91e3af135a4f4b5874… > ‌ In Scalar\_calc\_Tmunu.F90, in Scalar\_calc\_Tmunu, xx and rr should be private in OpenMP in grid loop. fixed in https://bitbucket.org/canuda/scalar/commits/44027452fc248b00e2e99b31d42cb23… -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2485/include-complex-a…

1 0

#2630: WaveMoL example computes energy from unsynchronized grid variables
by Roland Haas 16 Aug '22

16 Aug '22

#2630: WaveMoL example computes energy from unsynchronized grid variables Reporter: Roland Haas Status: open Milestone: Version: Type: bug Priority: major Component: Changes (by Roland Haas): status: open (was new) -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2630/wavemol-example-c…

1 0

#2630: WaveMoL example computes energy from unsynchronized grid variables
by Roland Haas 16 Aug '22

16 Aug '22

#2630: WaveMoL example computes energy from unsynchronized grid variables Reporter: Roland Haas Status: new Milestone: Version: Type: bug Priority: major Component: Comment (by Roland Haas): Please review. -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2630/wavemol-example-c…

1 0

#2630: WaveMoL example computes energy from unsynchronized grid variables
by Roland Haas 16 Aug '22

16 Aug '22

#2630: WaveMoL example computes energy from unsynchronized grid variables Reporter: Roland Haas Status: new Milestone: Version: Type: bug Priority: major Component: The WaveMoL example in CactusExamples computes the wave energy as `E = phi_t**2 + phi_i*phi_i` over the whole grid. However it does so on `MoL_PostStep` and ends up doing so before it applies boundary conditions and `SYNC` to the evolved variables \(`phi`, `phi_t` and `phi_i`\). Pull request [https://bitbucket.org/cactuscode/cactusexamples/pull-requests/3/wavemol-com… addresses this. -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2630/wavemol-example-c…

1 0

#2549: inlcude FLRWSolver in ET
by Roland Haas 15 Aug '22

15 Aug '22

#2549: inlcude FLRWSolver in ET Reporter: Roland Haas Status: new Milestone: ET_2022_05 Version: development version Type: enhancement Priority: major Component: Comment (by Roland Haas): Chi Tian agreed to provide science level input if needed. -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2549/inlcude-flrwsolve…

1 0

#2629: MoL_PseudoEvolution vs ANALYSIS
by Gabriele Bozzola 12 Aug '22

12 Aug '22

#2629: MoL_PseudoEvolution vs ANALYSIS Reporter: Gabriele Bozzola Status: new Milestone: Version: Type: bug Priority: major Component: Changes (by Gabriele Bozzola): Consider a thorn like `NPScalars` in the canuda arrangement. `NPScalars` computes the Newman-Penrose grid functions in the `CCTK_ANALYSIS` bin. The operation requires taking derivatives. Similarly, `LeanBSSNMoL` computes the constraints in the same bin \(this will likely be a regression I introduced in commit 80b6b1b\). My understanding is now that `CCTK_ANALYSIS` is not the right place where to do this because of interpolation/prolongation/notsurewhat. Is the fix to simply replace `CCTK_ANALYSIS` with `MoL_PseudoEvolution`? I would like to make sure that this is the case considering also parameters like `compute_every`. Since these are expensive diagnostics, it is best to compute them only when they are output. Do I get the correct constraints/scalars if I compute and output them only when all the refinement levels are synced? ‌ -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2629/mol_pseudoevoluti…

1 0

#2629: MoL_PseudoEvolution vs ANALYSIS
by Gabriele Bozzola 12 Aug '22

12 Aug '22

#2629: MoL_PseudoEvolution vs ANALYSIS Reporter: Gabriele Bozzola Status: new Milestone: Version: Type: bug Priority: major Component: Changes (by Gabriele Bozzola): Consider a thorn like `NPScalars` in the canuda arrangement. `NPScalars` computes the Newman-Penrose grid functions in the `CCTK_ANALYSIS` bin. The operation requires taking derivative. Similarly, `LeanBSSNMoL` computes the constraints in the same bin \(this will likely be a regression I introduced in commit 80b6b1b\). My understanding is now that `CCTK_ANALYSIS` is not the right place where to do this because of interpolation/prolongation/notsurewhat. Is the fix to simply replace `CCTK_ANALYSIS` with `MoL_PseudoEvolution`? I would like to make sure that this is the case considering also parameters like `compute_every`. Since these are expensive diagnostics, it is best to compute them only when they are output. Do I get the correct constraints/scalars if I compute and output them only when all the refinement levels are synced? ‌ -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2629/mol_pseudoevoluti…

1 0

#2629: MoL_PseudoEvolution vs ANALYSIS
by Gabriele Bozzola 12 Aug '22

12 Aug '22

#2629: MoL_PseudoEvolution vs ANALYSIS Reporter: Gabriele Bozzola Status: new Milestone: Version: Type: bug Priority: major Component: Consider a thorn like `NPScalars` in the canuda arrangement. `NPScalars` computes the Newman-Penrose grid functions in the `CCTK_ANALYSIS` bin. The operation requires taking derivative. Similarly, `LeanBSSNMoL` computes the constraints in the same bin \(this will likely be a regression I introduced in commit 80b6b1b\). My understanding is now that `CCTK_ANALYSIS` is not the right place where to do this because of interpolation/prolongation/notsurewhat. Is the fix to simply replace `CCTK_ANALYSIS` with `MoL_PseudoEvolution`? I would like to make sure that this is the considering also parameters like `compute_every`. Since these are expensive diagnostics, it is best to compute them only when they are output. Do I get the correct constraints/scalars if I compute and output them only when all the refinement levels are synced? ‌ -- Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2629/mol_pseudoevoluti…

1 0

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Trac