Remove solver_.parameters_ #668

montythind · 2024-09-24T22:41:45Z

This PR includes moving Parameters
absolute tolerance and relative tolearnce to state from solver

Note:
I have left
std::vector absolute_tolerance_;
double relative_tolerance_{ 1e-6 };

in rosenbrock_solver_prarameter.hpp as the CUDA tests are failing because tolerance doesnt get copied over to the device. Issue occurs in CudaRosenbrockSolver constructor when doing this->parameters_.absolute_tolerance_.size(); it comes out as zero

We can potential get rid of this line in solver_builder.inl. This is also a temporary. solution.
options.absolute_tolerance_ = state_parameters.absolute_tolerance_;

Along with that i have commented out unit tests till we don't have a solution how to deal with tolerance.

closes #621

montythind · 2024-09-24T22:45:37Z

I do see some calls where solver.solver_.parameters_ is not being modifiedbut rather the value is getting used. Example:

EXPECT_EQ(solver.solver_.parameters_.absolute_tolerance_[state.variable_map_["Ar"]], 1.0e-12);
std::vector atol = gpu_solver.solver_.parameters_.absolute_tolerance_;

I am unsure if we need to change these as well?

@K20shores @mattldawson

codecov-commenter · 2024-09-24T22:58:39Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 94.28%. Comparing base (25dee9b) to head (69fbc05).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #668      +/-   ##
==========================================
+ Coverage   93.97%   94.28%   +0.31%     
==========================================
  Files          61       61              
  Lines        4163     4179      +16     
==========================================
+ Hits         3912     3940      +28     
+ Misses        251      239      -12

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

K20shores

Right direction, but we still have some more work to do. Any mechanism-specific thing (like tolerances) need to be moved somewhere else. We should spend some time figuring out where they need to go. Also, the parameters need to be any set of paramters (rosenbrock, backward euler, etc.)

include/micm/solver/solver.hpp

sjsprecious

Thanks @montythind for working on this PR. I observed the following failures on Derecho:

         15 - cuda_rosenbrock (SEGFAULT)
         31 - rosenbrock (SEGFAULT)
         45 - regression_test_solve (NUMERICAL)
         46 - chapman_integration (SEGFAULT)
         47 - analytical_rosenbrock_integration (NUMERICAL)
         48 - analytical_backward_euler (NUMERICAL)
         49 - terminator (NUMERICAL)
         50 - integrated_reaction_rates (NUMERICAL)
         52 - README_example (NUMERICAL)
         53 - solve_results (NUMERICAL)
         54 - multiple_grid_cells (NUMERICAL)
         55 - rate_constants_no_user_defined_example_by_hand (NUMERICAL)
         56 - rate_constants_user_defined_example_by_hand (NUMERICAL)
         57 - solver_configuration (NUMERICAL)
         59 - rate_constants_no_user_defined_example_with_config (NUMERICAL)
         60 - rate_constants_user_defined_example_with_config (NUMERICAL)
         61 - carbon_bond_5_example (NUMERICAL)
         62 - chapman_example (NUMERICAL)
         63 - robertson_example (NUMERICAL)
         64 - ts1_example (NUMERICAL)

test/unit/cuda/solver/test_cuda_solver_builder.cpp

include/micm/solver/rosenbrock_solver_parameters.hpp

montythind · 2024-10-17T17:14:05Z

I am not sure what happened but after my latest pull and merge- lot of tests failed. I am looking what caused the test failures

sjsprecious · 2024-11-11T17:32:49Z

Interesting. I just ran the CUDA tests on my branch for the LU decomposition rename PR and it ran much faster.

Test project /glade/derecho/scratch/sunjian/micm/build
      Start  1: solver_config
 1/64 Test  #1: solver_config ........................................   Passed    0.68 sec
      Start  2: arrhenius_config
 2/64 Test  #2: arrhenius_config .....................................   Passed    0.70 sec
      Start  3: branched_config
 3/64 Test  #3: branched_config ......................................   Passed    0.80 sec
      Start  4: emission_config
 4/64 Test  #4: emission_config ......................................   Passed    0.28 sec
      Start  5: first_order_loss_config
 5/64 Test  #5: first_order_loss_config ..............................   Passed    0.40 sec
      Start  6: photolysis_config
 6/64 Test  #6: photolysis_config ....................................   Passed    0.67 sec
      Start  7: surface_config
 7/64 Test  #7: surface_config .......................................   Passed    0.42 sec
      Start  8: ternary_chemical_activation_config
 8/64 Test  #8: ternary_chemical_activation_config ...................   Passed    0.65 sec
      Start  9: troe_config
 9/64 Test  #9: troe_config ..........................................   Passed    0.33 sec
      Start 10: tunneling_config
10/64 Test #10: tunneling_config .....................................   Passed    0.23 sec
      Start 11: user_defined_config
11/64 Test #11: user_defined_config ..................................   Passed    0.39 sec
      Start 12: cuda_process_set
12/64 Test #12: cuda_process_set .....................................   Passed    6.24 sec
      Start 13: cuda_doolittle_lu_decomposition
13/64 Test #13: cuda_doolittle_lu_decomposition ......................   Passed    6.24 sec
      Start 14: cuda_linear_solver
14/64 Test #14: cuda_linear_solver ...................................   Passed    3.10 sec
      Start 15: cuda_rosenbrock
15/64 Test #15: cuda_rosenbrock ......................................   Passed    2.37 sec
      Start 16: cuda_solver_builder
16/64 Test #16: cuda_solver_builder ..................................   Passed    0.47 sec
      Start 17: cuda_dense_matrix
17/64 Test #17: cuda_dense_matrix ....................................   Passed    0.52 sec
      Start 18: cuda_sparse_matrix
18/64 Test #18: cuda_sparse_matrix ...................................   Passed    0.47 sec
      Start 19: process
19/64 Test #19: process ..............................................   Passed    0.05 sec
      Start 20: process_set
20/64 Test #20: process_set ..........................................   Passed    0.07 sec
      Start 21: arrhenius_rate_constant
21/64 Test #21: arrhenius_rate_constant ..............................   Passed    0.06 sec
      Start 22: branched_rate_constant
22/64 Test #22: branched_rate_constant ...............................   Passed    0.05 sec
      Start 23: surface_rate_constant
23/64 Test #23: surface_rate_constant ................................   Passed    0.06 sec
      Start 24: ternary_chemical_activation_rate_constant
24/64 Test #24: ternary_chemical_activation_rate_constant ............   Passed    0.05 sec
      Start 25: troe_rate_constant
25/64 Test #25: troe_rate_constant ...................................   Passed    0.03 sec
      Start 26: tunneling_rate_constant
26/64 Test #26: tunneling_rate_constant ..............................   Passed    0.10 sec
      Start 27: user_defined_rate_constant
27/64 Test #27: user_defined_rate_constant ...........................   Passed    0.03 sec
      Start 28: chapman_ode_solver
28/64 Test #28: chapman_ode_solver ...................................   Passed    0.10 sec
      Start 29: linear_solver
29/64 Test #29: linear_solver ........................................   Passed    0.13 sec
      Start 30: doolittle_lu_decomposition
30/64 Test #30: doolittle_lu_decomposition ...........................   Passed    0.10 sec
      Start 31: rosenbrock
31/64 Test #31: rosenbrock ...........................................   Passed    0.04 sec
      Start 32: state
32/64 Test #32: state ................................................   Passed    0.07 sec
      Start 33: backward_euler
33/64 Test #33: backward_euler .......................................   Passed    0.04 sec
      Start 34: solver_builder
34/64 Test #34: solver_builder .......................................   Passed    0.05 sec
      Start 35: phase
35/64 Test #35: phase ................................................   Passed    0.05 sec
      Start 36: species
36/64 Test #36: species ..............................................   Passed    0.10 sec
      Start 37: system
37/64 Test #37: system ...............................................   Passed    0.08 sec
      Start 38: matrix
38/64 Test #38: matrix ...............................................   Passed    0.07 sec
      Start 39: sparse_matrix_standard_ordering
39/64 Test #39: sparse_matrix_standard_ordering ......................   Passed    0.04 sec
      Start 40: sparse_matrix_vector_ordering
40/64 Test #40: sparse_matrix_vector_ordering ........................   Passed    0.05 sec
      Start 41: vector_matrix
41/64 Test #41: vector_matrix ........................................   Passed    0.04 sec
      Start 42: version
42/64 Test #42: version ..............................................   Passed    0.06 sec
      Start 43: regression_test_rosenbrock_p_force
43/64 Test #43: regression_test_rosenbrock_p_force ...................   Passed    0.03 sec
      Start 44: regression_test_rosenbrock_dforce_dy
44/64 Test #44: regression_test_rosenbrock_dforce_dy .................   Passed    0.05 sec
      Start 45: regression_test_solve
45/64 Test #45: regression_test_solve ................................   Passed    0.04 sec
      Start 46: chapman_integration
46/64 Test #46: chapman_integration ..................................   Passed    0.05 sec
      Start 47: analytical_rosenbrock_integration
47/64 Test #47: analytical_rosenbrock_integration ....................   Passed    3.14 sec
      Start 48: analytical_backward_euler
48/64 Test #48: analytical_backward_euler ............................   Passed    0.83 sec
      Start 49: terminator
49/64 Test #49: terminator ...........................................   Passed    0.18 sec
      Start 50: integrated_reaction_rates
50/64 Test #50: integrated_reaction_rates ............................   Passed    0.27 sec
      Start 51: cuda_analytical_rosenbrock_integration
51/64 Test #51: cuda_analytical_rosenbrock_integration ...............   Passed  204.09 sec
      Start 52: README_example
52/64 Test #52: README_example .......................................   Passed    0.25 sec
      Start 53: solve_results
53/64 Test #53: solve_results ........................................   Passed    0.04 sec
      Start 54: multiple_grid_cells
54/64 Test #54: multiple_grid_cells ..................................   Passed    0.08 sec
      Start 55: rate_constants_no_user_defined_example_by_hand
55/64 Test #55: rate_constants_no_user_defined_example_by_hand .......   Passed    0.03 sec
      Start 56: rate_constants_user_defined_example_by_hand
56/64 Test #56: rate_constants_user_defined_example_by_hand ..........   Passed    0.04 sec
      Start 57: solver_configuration
57/64 Test #57: solver_configuration .................................   Passed    0.25 sec
      Start 58: vectorized_matrix_solver
58/64 Test #58: vectorized_matrix_solver .............................   Passed    0.07 sec
      Start 59: rate_constants_no_user_defined_example_with_config
59/64 Test #59: rate_constants_no_user_defined_example_with_config ...   Passed    0.10 sec
      Start 60: rate_constants_user_defined_example_with_config
60/64 Test #60: rate_constants_user_defined_example_with_config ......   Passed    0.13 sec
      Start 61: carbon_bond_5_example
61/64 Test #61: carbon_bond_5_example ................................   Passed    0.16 sec
      Start 62: chapman_example
62/64 Test #62: chapman_example ......................................   Passed    0.06 sec
      Start 63: robertson_example
63/64 Test #63: robertson_example ....................................   Passed    0.08 sec
      Start 64: ts1_example
64/64 Test #64: ts1_example ..........................................   Passed    0.65 sec

Can you run it on a scratch directory in case it is an issue due to the filesystem for the home directory?

K20shores · 2024-11-11T17:41:01Z

Interesting. I just ran the CUDA tests on my branch for the LU decomposition rename PR and it ran much faster.

Test project /glade/derecho/scratch/sunjian/micm/build
      Start  1: solver_config
 1/64 Test  #1: solver_config ........................................   Passed    0.68 sec
      Start  2: arrhenius_config
 2/64 Test  #2: arrhenius_config .....................................   Passed    0.70 sec
      Start  3: branched_config
 3/64 Test  #3: branched_config ......................................   Passed    0.80 sec
      Start  4: emission_config
 4/64 Test  #4: emission_config ......................................   Passed    0.28 sec
      Start  5: first_order_loss_config
 5/64 Test  #5: first_order_loss_config ..............................   Passed    0.40 sec
      Start  6: photolysis_config
 6/64 Test  #6: photolysis_config ....................................   Passed    0.67 sec
      Start  7: surface_config
 7/64 Test  #7: surface_config .......................................   Passed    0.42 sec
      Start  8: ternary_chemical_activation_config
 8/64 Test  #8: ternary_chemical_activation_config ...................   Passed    0.65 sec
      Start  9: troe_config
 9/64 Test  #9: troe_config ..........................................   Passed    0.33 sec
      Start 10: tunneling_config
10/64 Test #10: tunneling_config .....................................   Passed    0.23 sec
      Start 11: user_defined_config
11/64 Test #11: user_defined_config ..................................   Passed    0.39 sec
      Start 12: cuda_process_set
12/64 Test #12: cuda_process_set .....................................   Passed    6.24 sec
      Start 13: cuda_doolittle_lu_decomposition
13/64 Test #13: cuda_doolittle_lu_decomposition ......................   Passed    6.24 sec
      Start 14: cuda_linear_solver
14/64 Test #14: cuda_linear_solver ...................................   Passed    3.10 sec
      Start 15: cuda_rosenbrock
15/64 Test #15: cuda_rosenbrock ......................................   Passed    2.37 sec
      Start 16: cuda_solver_builder
16/64 Test #16: cuda_solver_builder ..................................   Passed    0.47 sec
      Start 17: cuda_dense_matrix
17/64 Test #17: cuda_dense_matrix ....................................   Passed    0.52 sec
      Start 18: cuda_sparse_matrix
18/64 Test #18: cuda_sparse_matrix ...................................   Passed    0.47 sec
      Start 19: process
19/64 Test #19: process ..............................................   Passed    0.05 sec
      Start 20: process_set
20/64 Test #20: process_set ..........................................   Passed    0.07 sec
      Start 21: arrhenius_rate_constant
21/64 Test #21: arrhenius_rate_constant ..............................   Passed    0.06 sec
      Start 22: branched_rate_constant
22/64 Test #22: branched_rate_constant ...............................   Passed    0.05 sec
      Start 23: surface_rate_constant
23/64 Test #23: surface_rate_constant ................................   Passed    0.06 sec
      Start 24: ternary_chemical_activation_rate_constant
24/64 Test #24: ternary_chemical_activation_rate_constant ............   Passed    0.05 sec
      Start 25: troe_rate_constant
25/64 Test #25: troe_rate_constant ...................................   Passed    0.03 sec
      Start 26: tunneling_rate_constant
26/64 Test #26: tunneling_rate_constant ..............................   Passed    0.10 sec
      Start 27: user_defined_rate_constant
27/64 Test #27: user_defined_rate_constant ...........................   Passed    0.03 sec
      Start 28: chapman_ode_solver
28/64 Test #28: chapman_ode_solver ...................................   Passed    0.10 sec
      Start 29: linear_solver
29/64 Test #29: linear_solver ........................................   Passed    0.13 sec
      Start 30: doolittle_lu_decomposition
30/64 Test #30: doolittle_lu_decomposition ...........................   Passed    0.10 sec
      Start 31: rosenbrock
31/64 Test #31: rosenbrock ...........................................   Passed    0.04 sec
      Start 32: state
32/64 Test #32: state ................................................   Passed    0.07 sec
      Start 33: backward_euler
33/64 Test #33: backward_euler .......................................   Passed    0.04 sec
      Start 34: solver_builder
34/64 Test #34: solver_builder .......................................   Passed    0.05 sec
      Start 35: phase
35/64 Test #35: phase ................................................   Passed    0.05 sec
      Start 36: species
36/64 Test #36: species ..............................................   Passed    0.10 sec
      Start 37: system
37/64 Test #37: system ...............................................   Passed    0.08 sec
      Start 38: matrix
38/64 Test #38: matrix ...............................................   Passed    0.07 sec
      Start 39: sparse_matrix_standard_ordering
39/64 Test #39: sparse_matrix_standard_ordering ......................   Passed    0.04 sec
      Start 40: sparse_matrix_vector_ordering
40/64 Test #40: sparse_matrix_vector_ordering ........................   Passed    0.05 sec
      Start 41: vector_matrix
41/64 Test #41: vector_matrix ........................................   Passed    0.04 sec
      Start 42: version
42/64 Test #42: version ..............................................   Passed    0.06 sec
      Start 43: regression_test_rosenbrock_p_force
43/64 Test #43: regression_test_rosenbrock_p_force ...................   Passed    0.03 sec
      Start 44: regression_test_rosenbrock_dforce_dy
44/64 Test #44: regression_test_rosenbrock_dforce_dy .................   Passed    0.05 sec
      Start 45: regression_test_solve
45/64 Test #45: regression_test_solve ................................   Passed    0.04 sec
      Start 46: chapman_integration
46/64 Test #46: chapman_integration ..................................   Passed    0.05 sec
      Start 47: analytical_rosenbrock_integration
47/64 Test #47: analytical_rosenbrock_integration ....................   Passed    3.14 sec
      Start 48: analytical_backward_euler
48/64 Test #48: analytical_backward_euler ............................   Passed    0.83 sec
      Start 49: terminator
49/64 Test #49: terminator ...........................................   Passed    0.18 sec
      Start 50: integrated_reaction_rates
50/64 Test #50: integrated_reaction_rates ............................   Passed    0.27 sec
      Start 51: cuda_analytical_rosenbrock_integration
51/64 Test #51: cuda_analytical_rosenbrock_integration ...............   Passed  204.09 sec
      Start 52: README_example
52/64 Test #52: README_example .......................................   Passed    0.25 sec
      Start 53: solve_results
53/64 Test #53: solve_results ........................................   Passed    0.04 sec
      Start 54: multiple_grid_cells
54/64 Test #54: multiple_grid_cells ..................................   Passed    0.08 sec
      Start 55: rate_constants_no_user_defined_example_by_hand
55/64 Test #55: rate_constants_no_user_defined_example_by_hand .......   Passed    0.03 sec
      Start 56: rate_constants_user_defined_example_by_hand
56/64 Test #56: rate_constants_user_defined_example_by_hand ..........   Passed    0.04 sec
      Start 57: solver_configuration
57/64 Test #57: solver_configuration .................................   Passed    0.25 sec
      Start 58: vectorized_matrix_solver
58/64 Test #58: vectorized_matrix_solver .............................   Passed    0.07 sec
      Start 59: rate_constants_no_user_defined_example_with_config
59/64 Test #59: rate_constants_no_user_defined_example_with_config ...   Passed    0.10 sec
      Start 60: rate_constants_user_defined_example_with_config
60/64 Test #60: rate_constants_user_defined_example_with_config ......   Passed    0.13 sec
      Start 61: carbon_bond_5_example
61/64 Test #61: carbon_bond_5_example ................................   Passed    0.16 sec
      Start 62: chapman_example
62/64 Test #62: chapman_example ......................................   Passed    0.06 sec
      Start 63: robertson_example
63/64 Test #63: robertson_example ....................................   Passed    0.08 sec
      Start 64: ts1_example
64/64 Test #64: ts1_example ..........................................   Passed    0.65 sec

Can you run it on a scratch directory in case it is an issue due to the filesystem for the home directory?

Oh man. That's drastically longer. I will try scratch and see what happens

K20shores · 2024-11-11T20:43:56Z

@sjsprecious I guess it was a scratch issue?

Running tests...
Test project /glade/derecho/scratch/kshores/micm/build
      Start  1: solver_config
 1/64 Test  #1: solver_config ........................................   Passed    0.19 sec
      Start  2: arrhenius_config
 2/64 Test  #2: arrhenius_config .....................................   Passed    0.13 sec
      Start  3: branched_config
 3/64 Test  #3: branched_config ......................................   Passed    0.11 sec
      Start  4: emission_config
 4/64 Test  #4: emission_config ......................................   Passed    0.09 sec
      Start  5: first_order_loss_config
 5/64 Test  #5: first_order_loss_config ..............................   Passed    0.05 sec
      Start  6: photolysis_config
 6/64 Test  #6: photolysis_config ....................................   Passed    0.10 sec
      Start  7: surface_config
 7/64 Test  #7: surface_config .......................................   Passed    0.10 sec
      Start  8: ternary_chemical_activation_config
 8/64 Test  #8: ternary_chemical_activation_config ...................   Passed    0.06 sec
      Start  9: troe_config
 9/64 Test  #9: troe_config ..........................................   Passed    0.05 sec
      Start 10: tunneling_config
10/64 Test #10: tunneling_config .....................................   Passed    0.07 sec
      Start 11: user_defined_config
11/64 Test #11: user_defined_config ..................................   Passed    0.10 sec
      Start 12: cuda_process_set
12/64 Test #12: cuda_process_set .....................................   Passed    5.88 sec
      Start 13: cuda_lu_decomposition
13/64 Test #13: cuda_lu_decomposition ................................   Passed    6.23 sec
      Start 14: cuda_linear_solver
14/64 Test #14: cuda_linear_solver ...................................   Passed    3.23 sec
      Start 15: cuda_rosenbrock
15/64 Test #15: cuda_rosenbrock ......................................   Passed    2.45 sec
      Start 16: cuda_solver_builder
16/64 Test #16: cuda_solver_builder ..................................   Passed    0.45 sec
      Start 17: cuda_dense_matrix
17/64 Test #17: cuda_dense_matrix ....................................   Passed    0.46 sec
      Start 18: cuda_sparse_matrix
18/64 Test #18: cuda_sparse_matrix ...................................   Passed    0.53 sec
      Start 19: process
19/64 Test #19: process ..............................................   Passed    0.05 sec
      Start 20: process_set
20/64 Test #20: process_set ..........................................   Passed    0.08 sec
      Start 21: arrhenius_rate_constant
21/64 Test #21: arrhenius_rate_constant ..............................   Passed    0.05 sec
      Start 22: branched_rate_constant
22/64 Test #22: branched_rate_constant ...............................   Passed    0.06 sec
      Start 23: surface_rate_constant
23/64 Test #23: surface_rate_constant ................................   Passed    0.04 sec
      Start 24: ternary_chemical_activation_rate_constant
24/64 Test #24: ternary_chemical_activation_rate_constant ............   Passed    0.05 sec
      Start 25: troe_rate_constant
25/64 Test #25: troe_rate_constant ...................................   Passed    0.06 sec
      Start 26: tunneling_rate_constant
26/64 Test #26: tunneling_rate_constant ..............................   Passed    0.04 sec
      Start 27: user_defined_rate_constant
27/64 Test #27: user_defined_rate_constant ...........................   Passed    0.05 sec
      Start 28: chapman_ode_solver
28/64 Test #28: chapman_ode_solver ...................................   Passed    0.09 sec
      Start 29: linear_solver
29/64 Test #29: linear_solver ........................................   Passed    0.14 sec
      Start 30: lu_decomposition
30/64 Test #30: lu_decomposition .....................................   Passed    0.06 sec
      Start 31: rosenbrock
31/64 Test #31: rosenbrock ...........................................   Passed    0.05 sec
      Start 32: state
32/64 Test #32: state ................................................   Passed    0.05 sec
      Start 33: backward_euler
33/64 Test #33: backward_euler .......................................   Passed    0.06 sec
      Start 34: solver_builder
34/64 Test #34: solver_builder .......................................   Passed    0.04 sec
      Start 35: phase
35/64 Test #35: phase ................................................   Passed    0.05 sec
      Start 36: species
36/64 Test #36: species ..............................................   Passed    0.07 sec
      Start 37: system
37/64 Test #37: system ...............................................   Passed    0.04 sec
      Start 38: matrix
38/64 Test #38: matrix ...............................................   Passed    0.07 sec
      Start 39: sparse_matrix_standard_ordering
39/64 Test #39: sparse_matrix_standard_ordering ......................   Passed    0.10 sec
      Start 40: sparse_matrix_vector_ordering
40/64 Test #40: sparse_matrix_vector_ordering ........................   Passed    0.05 sec
      Start 41: vector_matrix
41/64 Test #41: vector_matrix ........................................   Passed    0.05 sec
      Start 42: version
42/64 Test #42: version ..............................................   Passed    0.06 sec
      Start 43: regression_test_rosenbrock_p_force
43/64 Test #43: regression_test_rosenbrock_p_force ...................   Passed    0.05 sec
      Start 44: regression_test_rosenbrock_dforce_dy
44/64 Test #44: regression_test_rosenbrock_dforce_dy .................   Passed    0.07 sec
      Start 45: regression_test_solve
45/64 Test #45: regression_test_solve ................................   Passed    0.05 sec
      Start 46: chapman_integration
46/64 Test #46: chapman_integration ..................................   Passed    0.06 sec
      Start 47: analytical_rosenbrock_integration
47/64 Test #47: analytical_rosenbrock_integration ....................   Passed    2.94 sec
      Start 48: analytical_backward_euler
48/64 Test #48: analytical_backward_euler ............................   Passed    0.77 sec
      Start 49: terminator
49/64 Test #49: terminator ...........................................   Passed    0.22 sec
      Start 50: integrated_reaction_rates
50/64 Test #50: integrated_reaction_rates ............................   Passed    0.04 sec
      Start 51: cuda_analytical_rosenbrock_integration
51/64 Test #51: cuda_analytical_rosenbrock_integration ...............   Passed  241.47 sec
      Start 52: README_example
52/64 Test #52: README_example .......................................   Passed    0.06 sec
      Start 53: solve_results
53/64 Test #53: solve_results ........................................   Passed    0.04 sec
      Start 54: multiple_grid_cells
54/64 Test #54: multiple_grid_cells ..................................   Passed    0.07 sec
      Start 55: rate_constants_no_user_defined_example_by_hand
55/64 Test #55: rate_constants_no_user_defined_example_by_hand .......   Passed    0.04 sec
      Start 56: rate_constants_user_defined_example_by_hand
56/64 Test #56: rate_constants_user_defined_example_by_hand ..........   Passed    0.06 sec
      Start 57: solver_configuration
57/64 Test #57: solver_configuration .................................   Passed    0.10 sec
      Start 58: vectorized_matrix_solver
58/64 Test #58: vectorized_matrix_solver .............................   Passed    0.05 sec
      Start 59: rate_constants_no_user_defined_example_with_config
59/64 Test #59: rate_constants_no_user_defined_example_with_config ...   Passed    0.05 sec
      Start 60: rate_constants_user_defined_example_with_config
60/64 Test #60: rate_constants_user_defined_example_with_config ......   Passed    0.06 sec
      Start 61: carbon_bond_5_example
61/64 Test #61: carbon_bond_5_example ................................   Passed    0.14 sec
      Start 62: chapman_example
62/64 Test #62: chapman_example ......................................   Passed    0.01 sec
      Start 63: robertson_example
63/64 Test #63: robertson_example ....................................   Passed    0.01 sec
      Start 64: ts1_example
64/64 Test #64: ts1_example ..........................................   Passed    0.61 sec

sjsprecious · 2024-11-11T21:12:46Z

@K20shores The new result is consistent with mine now. I noticed that you were running the CUDA tests on your home directory on Derecho and I think that was the problem.

…allowSolversToModify_621

include/micm/configure/solver_config.hpp

include/micm/solver/solver_builder.hpp

… into allowSolversToModify_621

sjsprecious

Thanks @K20shores and @montythind for working on this and addressing all my comments. It looks great to me and I could confirm that all the CPU & GPU tests pass on Derecho.

mattldawson

Looks great! Just one minor question, but nothing that should hold up merging

mattldawson · 2025-01-09T22:23:23Z

include/micm/solver/solver.hpp

+      solver_.parameters_ = params;
+      return solver_.Solve(time_step, state);


Is it necessary to have parameters_ be a data member of solver_ still? Could we just pass the parameters to the solve function as an argument by reference and avoid the copy?

boulderdaze

Looks good! I just have a couple questions and suggestions.

boulderdaze · 2025-01-09T22:30:13Z

include/micm/configure/solver_config.hpp

    {
    }


I'd suggest initializing the struct variable in the other two constructors that don’t pass relative tolerance as a function argument.

SolverParameters(System&& system, std::vector<Process>&& processes, RosenbrockSolverParameters&& parameters) : system_(system), processes_(processes), parameters_(parameters) { relative_tolerance_ = 0.0; }

boulderdaze · 2025-01-09T22:37:08Z

include/micm/configure/solver_config.hpp

@@ -959,7 +968,7 @@ namespace micm
    SolverParameters GetSolverParams()
    {
      return SolverParameters(
-          std::move(System(this->gas_phase_, this->phases_)), std::move(this->processes_), std::move(this->parameters_));
+          std::move(System(this->gas_phase_, this->phases_)), std::move(this->processes_), std::move(this->parameters_), std::move(this->relative_tolerance_));


Could you remove std::move? It doesn’t match the constructor and I think there's no real need for that for a double.

this->relative_tolerance_;

boulderdaze · 2025-01-09T22:48:36Z

include/micm/solver/solver.hpp

@@ -65,6 +65,13 @@ namespace micm
      return result;
    }

+    // Overloaded Solve function to change parameters
+    SolverResult Solve(double time_step, StatePolicy& state, SolverParametersType& params)


Can they be passed as const ref since the objects are not modified?

SolverResult Solve(double time_step, const StatePolicy& state, const SolverParametersType& params)

boulderdaze · 2025-01-09T22:50:37Z

include/micm/solver/solver_builder.inl

+
+
+    this->SetAbsoluteTolerances(state_parameters.absolute_tolerance_, species_map);


Maybe this additional new line isn't necessary?

draft of 621

263af07

K20shores requested changes Sep 25, 2024

View reviewed changes

include/micm/solver/solver.hpp Outdated Show resolved Hide resolved

montythind added 2 commits September 30, 2024 13:11

second draft

da7cd22

third draft

32a16e2

montythind force-pushed the allowSolversToModify_621 branch from d059401 to 32a16e2 Compare October 1, 2024 21:07

montythind added 10 commits October 1, 2024 16:20

darft fourth

f125b16

Merge branch 'main' into allowSolversToModify_621

c580d79

test fixed

03a8105

fix cuda test

0904e97

fix cuda tests

f4cf73a

fix again

d67add6

cuda test fix

ecb857a

removed params

fefab05

added tolerance back

c481a0b

fix cuda tests

fdb1fb6

montythind changed the title ~~Draft PR: Remove solver_.parameters_~~ Remove solver_.parameters_ Oct 16, 2024

montythind requested a review from K20shores October 16, 2024 20:28

K20shores requested review from sjsprecious, mattldawson, boulderdaze and mwaxmonsky October 17, 2024 15:52

Merge branch 'main' into allowSolversToModify_621

1bc0df6

sjsprecious requested changes Oct 17, 2024

View reviewed changes

test/unit/cuda/solver/test_cuda_solver_builder.cpp Show resolved Hide resolved

include/micm/solver/rosenbrock_solver_parameters.hpp Outdated Show resolved Hide resolved

K20shores mentioned this pull request Oct 22, 2024

Jacobian diagonal elements should be moved to the state #681

Open

K20shores added 3 commits November 4, 2024 10:05

trying to copy tolerances from state

9f4b7b3

updating usage of absolute tolerance

4c84691

removed compile errors

1b67f79

added tests

6112c2c

montythind requested a review from sjsprecious November 13, 2024 19:31

sjsprecious mentioned this pull request Nov 13, 2024

Add a StateParameters data member in the State class #687

Open

sjsprecious added 4 commits November 15, 2024 15:21

move the copy of absolute tolerance into the cuda state constructor

bd1829e

Merge branch 'allowSolversToModify_621' of github.com:NCAR/micm into …

130b83c

…allowSolversToModify_621

fix broken CI test

5a612d6

uncomment a GPU test

bacb558

sjsprecious reviewed Nov 19, 2024

View reviewed changes

include/micm/configure/solver_config.hpp Show resolved Hide resolved

include/micm/solver/solver_builder.hpp Outdated Show resolved Hide resolved

K20shores and others added 14 commits January 8, 2025 11:49

handling merge

009f6ae

removing more template parameters

543d57a

removing template parameters

316d005

removing template parameters

a450599

removing another

d00baca

adding typename

ae93072

fixing merge conflict in cuda files

8f83390

trying to get cuda to compile

b63e8c7

getting cuda to compile

c6e89ab

merging main

eceb9a8

merging again

10b8ab0

removing variable

7fb1560

removing getter

d2770cf

Merge branch 'allowSolversToModify_621' of https://github.com/NCAR/micm…

69fbc05

… into allowSolversToModify_621

K20shores requested a review from sjsprecious January 9, 2025 20:51

sjsprecious approved these changes Jan 9, 2025

View reviewed changes

mattldawson approved these changes Jan 9, 2025

View reviewed changes

boulderdaze requested changes Jan 9, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Remove solver_.parameters_ #668

Remove solver_.parameters_ #668

montythind commented Sep 24, 2024 •

edited

Loading

montythind commented Sep 24, 2024

codecov-commenter commented Sep 24, 2024 •

edited

Loading

K20shores left a comment

sjsprecious left a comment

montythind commented Oct 17, 2024

sjsprecious commented Nov 11, 2024

K20shores commented Nov 11, 2024

K20shores commented Nov 11, 2024

sjsprecious commented Nov 11, 2024

sjsprecious left a comment

mattldawson left a comment

mattldawson Jan 9, 2025

boulderdaze left a comment

boulderdaze Jan 9, 2025

boulderdaze Jan 9, 2025

boulderdaze Jan 9, 2025

boulderdaze Jan 9, 2025

		solver_.parameters_ = params;
		return solver_.Solve(time_step, state);



		this->SetAbsoluteTolerances(state_parameters.absolute_tolerance_, species_map);

Remove solver_.parameters_ #668

Are you sure you want to change the base?

Remove solver_.parameters_ #668

Conversation

montythind commented Sep 24, 2024 • edited Loading

montythind commented Sep 24, 2024

codecov-commenter commented Sep 24, 2024 • edited Loading

Codecov Report

K20shores left a comment

Choose a reason for hiding this comment

sjsprecious left a comment

Choose a reason for hiding this comment

montythind commented Oct 17, 2024

sjsprecious commented Nov 11, 2024

K20shores commented Nov 11, 2024

K20shores commented Nov 11, 2024

sjsprecious commented Nov 11, 2024

sjsprecious left a comment

Choose a reason for hiding this comment

mattldawson left a comment

Choose a reason for hiding this comment

mattldawson Jan 9, 2025

Choose a reason for hiding this comment

boulderdaze left a comment

Choose a reason for hiding this comment

boulderdaze Jan 9, 2025

Choose a reason for hiding this comment

boulderdaze Jan 9, 2025

Choose a reason for hiding this comment

boulderdaze Jan 9, 2025

Choose a reason for hiding this comment

boulderdaze Jan 9, 2025

Choose a reason for hiding this comment

montythind commented Sep 24, 2024 •

edited

Loading

codecov-commenter commented Sep 24, 2024 •

edited

Loading