Release 2024 #44
Open
Release 2024 #44
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* extend kitrt script * added new neural models * extend to m3 models * added M3_2D models * added M2 and M4 models * configure ml optimizer for high order models * added configuration for high order models * add option for rotation postprcessing in neural networks. remove unneccessary python scripts * started rotational symmetric postprocessing * added rotational invariance postprocessing for m2 and m1 models * fix post merge bugs * created hohlraum mesh * add hohlraum tes case * add hohlraum test case * add hohlraum cfg filees * fixed hohlraum testcase * add hohlraum cfg files * changed hohlraum cfg * changed hohlraum testcase to isotropic inflow source boundary condition * added ghost cell bonudary for hohlraum testcase * update readme and linesource mesh creator * added proper scaling for linesource reference solution * regularized newton debugging
* added reduced optimizer for sampling * remove old debugging comments
* added new ansatz * debug new ansatz * branch should be abandoned here * debug new ansatz * fix scaling error new ansatz * fix config errors * temporarily fixed dynamic ansatz rotation bug * fix inheritance error for new ansatz * mesh acceleration * add structured hohlraum
* mesh acceleration * added floor value for starmap moment methods * enable accelleration * delete minimum value starmap
* added new ansatz * debug new ansatz * branch should be abandoned here * debug new ansatz * fix scaling error new ansatz * fix config errors * temporarily fixed dynamic ansatz rotation bug * fix inheritance error for new ansatz * mesh acceleration * add structured hohlraum * added spherical harmonics basis generator * parallelize datagen * added flag for dynamic ansatz * added rotated monomial basis * fixed compilation errors * fixed rotated basis linesource * added rotated hohlraum testcase * code cleanup * restructure old files * added regularized monomial m1 networks * added spherical harmonics inference and started tests on rotation * added sh models * added rotated sh networks * beautifier * added rotated sh 2d and 3d models. EXPERIMENTALLY * re-add mirroring * did dynamic ansatz after rotation * added symmetry correction * fixed 2d sh projection * added 2d sh to datagenerator * remove rotation from data-gen * update 2d mn solver * delete redundant logs * add rotation to 2d data-gen * removed debugging output * parallelize sampler again * added 2d sh rotator up do degree 4 * fixed rotation matrix * test run for sh2d rotated m1 closure * fix compile error * attempt to fix vector size bug * add option to turn of rotation * change m1_2d model * change m1_2d models * delete tf history files * bugfix: u_2 was manually set to 0 in 2d sh network evaluation - removed line * bugfix: dynamic closure was missing in non-rotated 2d monomial and sh NN closures. Removed rotation correction for m1 monomial NN closure * bugfix: false naming of 2d m1 sh model * bugfix: vector length for rotated sh neural network closure * renamed some vectors * fix: formatting, BUGFIX: rotation angle * bugfix: made several function arguments in optimizer const for type safety * enhance: made rotated datagen mirrored * ongoing: fix rotated mn bug for spherical harmonics * bufgfix: Rotated m1 works now * enhance: rotated m1 models added * bugfix: removed column deletion for higher order rotated models * formatting * modelchange: m1_2d_rot * change: dockerfile * enhance: added sh m2 models. change: newton optimizer stopping criterion to u recons. change: mn_normalized_ml solver * change: default values for newton * change: changed error measure of neural cells to relative norm error * enhance: add M3 spherical harmonics models * added measure for inaccuraate nn cells --------- Co-authored-by: Steffen <[email protected]>
* added new ansatz * debug new ansatz * branch should be abandoned here * debug new ansatz * fix scaling error new ansatz * fix config errors * temporarily fixed dynamic ansatz rotation bug * fix inheritance error for new ansatz * mesh acceleration * add structured hohlraum * added spherical harmonics basis generator * parallelize datagen * added flag for dynamic ansatz * added rotated monomial basis * fixed compilation errors * fixed rotated basis linesource * added rotated hohlraum testcase * code cleanup * restructure old files * added regularized monomial m1 networks * added spherical harmonics inference and started tests on rotation * added sh models * added rotated sh networks * beautifier * added rotated sh 2d and 3d models. EXPERIMENTALLY * re-add mirroring * did dynamic ansatz after rotation * added symmetry correction * fixed 2d sh projection * added 2d sh to datagenerator * remove rotation from data-gen * update 2d mn solver * delete redundant logs * add rotation to 2d data-gen * removed debugging output * parallelize sampler again * added 2d sh rotator up do degree 4 * fixed rotation matrix * test run for sh2d rotated m1 closure * fix compile error * attempt to fix vector size bug * add option to turn of rotation * change m1_2d model * change m1_2d models * delete tf history files * bugfix: u_2 was manually set to 0 in 2d sh network evaluation - removed line * bugfix: dynamic closure was missing in non-rotated 2d monomial and sh NN closures. Removed rotation correction for m1 monomial NN closure * bugfix: false naming of 2d m1 sh model * bugfix: vector length for rotated sh neural network closure * renamed some vectors * fix: formatting, BUGFIX: rotation angle * bugfix: made several function arguments in optimizer const for type safety * enhance: made rotated datagen mirrored * ongoing: fix rotated mn bug for spherical harmonics * bufgfix: Rotated m1 works now * enhance: rotated m1 models added * bugfix: removed column deletion for higher order rotated models * formatting * modelchange: m1_2d_rot * change: dockerfile * enhance: added sh m2 models. change: newton optimizer stopping criterion to u recons. change: mn_normalized_ml solver * change: default values for newton * change: changed error measure of neural cells to relative norm error * enhance: add M3 spherical harmonics models * added measure for inaccuraate nn cells * enabled M3 rotations --------- Co-authored-by: Steffen <[email protected]>
* added symmetric hohlraum. Changed the way neumann cells work. ALl old test cases are now broken * created lattice test case and mesh generator * startet new scalar outputs * first QOIs work * finished lattice QOI * added hohlraum qois * fixed small bug with openMP * fix bug with probing moments * project restructrue * debugging * started variation QOIs * added variation QOI * startet incorporation of the UM-Bridge framework * setup um-bridge test case lattice - errror with network connection * added design parameters for the lattice test case * refactoring of problem specific QOIs. put to the corresponding problem class * included pointer to quadrature class to problem. WIll help reduce the memory footprint of the solver * made all qois and design parameters part of the corresponding problem class * fixed compiler warning about included map * fix small bug * fixed ninja compile bug * started hohlraum probing qoi * fixed type conversion bug * fixed compile problems. Added sampling line hohlraum test case * added line variance * added green line scalar output * segfault somewhere in hohlraum * resordered meshing scripts. startet refined mesh for lattice * mesh refinement for lattice * updated requirements * fixed computation of the characteristic length in 2d * mesh io verbose * mesh io verbose * added folder for mesh files geo * cascading file lattice * repaired lattice geo file * fixed geo file * adapted cascade file * added geo file for rectangular meshes * segfault introduced * fixed quad grid * fixed segfault in hohlraum, linesource and checkerboard * added singularity container support * added lattice benchmark * small change in lattice testcase file * added solver output * made progression for transfinite elements in lattice test case variable * added lattice meshing command * removed MPI and Python dependencies from KiT-RT * added MPI singularity def file * updated geo file * new example file structure * changed mesh size to minimal size for example * small speedup of mesh connectivity * mesh has now access to config. io writer for connectiivty added * started to add connectivity loader * added mesh connectivity reader from file for faster recomputation * increased cfl number * update mn and pn solver to new, more memory efficient phhysics implementation * fixed bug in max absorption computation * added config files * updated meshing scripts * updated meshing scripts * updated runscript * fixed outpout of logfiles. part 1 * start debugging logger * fixed logger filename bug * added cfg file * update n80 meshfile to geo progression 0.5 * fixed bug in output verbosity * added output verbosity * fixed scatter xs bug * removed some compiler warnings. fixed test cases for new ghost cell implementation * fixed several compile warning about unused variables. adapted reference files for validationt test to new boundary conditions WITHOUT mesh connectivity file. There is a BUG in the connectivity reader or writer * fixed all test cases locally. added unit test for mesh connectivity file reader * fixed a double defnition bug * fixed mn test case * fixed screen output test case * changed workflow file * small fix * small fix * revert changes to worklow * revert change in worklof yaml * fix csv output formatting * fix output validation test for new formatting * changed license * fix * upped writeout precision * upped output precifion for logger * change ?? * loop unrolling sn solver * added hpc examples sym_hohlraum * added sym_hohlraum cfg * changed gitignore * fix segfault in 2d flux * temporarily changed mesh resolution * added quarterhohlraum * added quarterhohlraum cfg and mesh files * added quarter hohlraum for high res simulation speedup * added meshing functionality * added quarter hohlraum * reflecting boundary conditions for quarter hohlraum * qois for quarter hohlraum * removed cout from main.cpp * fix unit tests * started hohlraum design params * accelerated sn solver by 25% through memory access optimization in Flux update (#45) * added half lattice benchmark * added half lattice bench * updates upwind flux * fix logger unit test * Fixed Bug in 2nd order HPC SN solver (#47) * fix for unit tests? * started hpc sn solver * continue mem opt * continue hpc sn solver * continue hpc solver * order 1 solver accellerated' * lattice qois updated * clean up of hpc examples * started 2nd order solver * added half lattice and 2nd order * 2nd order solver ready with venkat limiter * fix output bug * added 2nd order in time (inefficiently) * bugfixing 2nd order * changed initial condiiton to 0 * arrg * linesource 1st order paralelized * compartmentalize fvm update and fluxes (10% performance loss) * fixed lattice * started 2nd order flux * restore bs limiter * fixed 2nd order lattice * temp comment out hohlraum specific stuff * fixed hohlraum 2nd order with all QOIS --------- Co-authored-by: Steffen Schotthöfer <[email protected]> * Memory optimization solver (#46) * fix for unit tests? * started hpc sn solver * continue mem opt * continue hpc sn solver * continue hpc solver * order 1 solver accellerated' * lattice qois updated * clean up of hpc examples * started 2nd order solver * added half lattice and 2nd order * 2nd order solver ready with venkat limiter * fix output bug * added 2nd order in time (inefficiently) * bugfixing 2nd order * changed initial condiiton to 0 --------- Co-authored-by: Steffen Schotthöfer <[email protected]> * fixed small debug verbosity bugs * fix unit tests * fix screen output test * fixed screen output test - accounting for extra time step to fill up to t_f * fixes csv logger test case * fixed logger output now? * another logger fix * added container builder * remove debug output * rename sy hohlruam cfg * prepared for 2nd order in time * updated hohlraum qois * added singularity install script kitrt * changed flux limiter to venkatakrishnan * quarter hohlraum now works * fix segfaults * fix race condition * added variable green region to hozhlraum * variable red absorption region * fixed bug in boundary condition quarter hohlraum * fixed vtk memory leak * updated hohlraum design params * update options * comment unused var out * added lattice qois according to testbed v2 * updated quarter hohlraum test case * bugfix * parallel conmat (experimental) * parallel mesh connectivity gives segfault * mesh connectivity accelerated - experimental * accelerated solver startup * set forced connectivity wirte as an option * does not write connectivity file anymore when forced connectivity is active * fixed bug in cell of coord identification * added moment output for hpc sn solver * fixed 2nd order solver * fixed bug in absorption calculation of the hohlraum center capsule * added total absorption as output field * fixed hohrlaum absorption coefficient * Mpi sn solver (#49) * remove debugging output * fixed mpi bugs * fixed bug in logging mechanism * hohlraum outputs mpi parallelized * fixed output bug in hohlraum absorption * fix green domain * test ball probing moments * fixed parameter issue with hohlraum * fixed hohlraum parameters. Consitent with original crocett paper from 2019 * fixed issue when cell length is larger than ball * fixed indexing bug * fix reference file for screen output * remove meshing tests - not needed --------- Co-authored-by: Steffen Schotthoefer Co-authored-by: Steffen Schotthoefer
ScSteffen
requested review from
piastammer and
JonasKu
and removed request for
piastammer
piastammer
approved these changes
Sep 16, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed Changes
Give a brief overview of your contribution here in a few sentences.
Related Work
Resolve any issues (bug fix or feature request), note any related PRs, or mention interactions with the work of others, if any.
PR Checklist
Put an X by all that apply. You can fill this out after submitting the PR.