Error when creating 2d burgers problem

[jacktang]

Hello,

I want to test some instationary problems and find the burgers problems(1d/2d) are the built-in instationary problems. So I want to create the instance and visualize it. Here is the code

# p = burgers_problem()
p = burgers_problem_2d()

fom, fom_data = discretize_instationary_cg(p, time_stepper=1, diameter=1./5.)

print(fom_data['grid'])

U = fom.solve([1.0, 0.1, 0.3, 0.1, 0.2, 1.0])
fom.visualize(U, title='Solution')

And it throws NotImplementedError . Here is the exception traceback:

NotImplementedError                       Traceback (most recent call last)
Cell In[7], line 4
      1 p = burgers_problem_2d()
      2 # p = burgers_problem()
----> 4 fom, fom_data = discretize_instationary_cg(p, time_stepper=1, diameter=1.[/](https://file+.vscode-resource.vscode-cdn.net/)5.)
      6 print(fom_data['grid'])
      8 U = fom.solve([1.0, 0.1, 0.3, 0.1, 0.2, 1.0])

File [~/opt/miniconda3/envs/pymor/lib/python3.10/site-packages/pymor/discretizers/builtin/cg.py:1286](https://file+.vscode-resource.vscode-cdn.net/Users/apple/Codes/caebigdata/rom/pymor/~/opt/miniconda3/envs/pymor/lib/python3.10/site-packages/pymor/discretizers/builtin/cg.py:1286), in discretize_instationary_cg(analytical_problem, diameter, domain_discretizer, grid_type, grid, boundary_info, num_values, time_stepper, nt, preassemble)
   1281 if p.stationary_part.dirichlet_data is not None and 't' in p.stationary_part.dirichlet_data.parameters:
   1282     # we choose both mass and operator to be invertible.
   1283     # this leads to wrong results when the dirichlet values depend on time.
   1284     raise NotImplementedError('Time-dependent Dirichlet values not supported.')
-> 1286 m, data = discretize_stationary_cg(p.stationary_part, diameter=diameter, domain_discretizer=domain_discretizer,
   1287                                    grid_type=grid_type, grid=grid, boundary_info=boundary_info)
   1289 if p.initial_data.parametric:
   1290     I = InterpolationOperator(data['grid'], p.initial_data)

File [~/opt/miniconda3/envs/pymor/lib/python3.10/site-packages/pymor/discretizers/builtin/cg.py:1000](https://file+.vscode-resource.vscode-cdn.net/Users/apple/Codes/caebigdata/rom/pymor/~/opt/miniconda3/envs/pymor/lib/python3.10/site-packages/pymor/discretizers/builtin/cg.py:1000), in discretize_stationary_cg(analytical_problem, diameter, domain_discretizer, grid_type, grid, boundary_info, preassemble, mu_energy_product)
    993 p = analytical_problem
    995 if not (p.nonlinear_advection
    996         == p.nonlinear_advection_derivative
    997         == p.nonlinear_reaction
    998         == p.nonlinear_reaction_derivative
...
-> 1000     raise NotImplementedError
   1002 if grid is None:
   1003     domain_discretizer = domain_discretizer or discretize_domain_default

The problem definition is not completed?

[sdrave]

Hi @jacktang, non-linear advection is only implemented for finite volumes. Using discretize_instationary_fv should work.

[HenKlei]

The following works:

from pymor.basic import *

p = burgers_problem_2d()

fom, fom_data = discretize_instationary_fv(p, diameter=1./20., nt=100)

print(fom_data['grid'])

for mu in [0.1, 0.2, 0.3, 1.0]:
    U = fom.solve(mu)
    fom.visualize(U, title=f'Solution exponent={mu}')