{ "id": "single-software-execution--cli--fenics--task-06--ubuntu", "snapshot": "FEniCS", "instruction": "Treat a legacy FEniCS/DOLFIN nonlinear Poisson-equation template `/home/user/Desktop/nonlinear_base.py` as the starting point. Run with the legacy `dolfin` API used by the template (for example, FEniCS/DOLFIN 2019.1.0), not DOLFINx. On this machine, run legacy FEniCS with `/home/user/miniconda3/bin/conda run --no-capture-output --prefix /home/user/miniconda3/envs/fenics-legacy python` when working on `nonlinear_base.py`.\n\nComplete the following: 1. Modify and generate `/home/user/Desktop/nonlinear.py`: - Define nonlinear residual: `F = (1 + u**2)*inner(grad(u), grad(v))*dx - f*v*dx`, with `f=Constant(1.0)` - Define its Jacobian with `J = derivative(F, u, du)`, where `du = TrialFunction(V)`, and use `NonlinearVariationalProblem(F, u, bc, J)` - Set: `solver.parameters['newton_solver']['relative_tolerance'] = 1e-6` - Solve 2. Extract the value at center `(0.5, 0.5)` and the global maximum; for this symmetric mesh the center is the maximum up to numerical tolerance - Save to `/home/user/Desktop/summary.txt` in this format: `center_value, max_value`", "source": "https://olddocs.fenicsproject.org/dolfin/2019.1.0/python/", "config": [ { "type": "upload_file", "parameters": { "files": [ { "local_path": "task-06/init_file/nonlinear_base.py", "path": "/home/user/Desktop/nonlinear_base.py" } ] } } ], "trajectory": "trajectories/", "related_apps": [ "fenics" ], "artifact_paths": [ "/home/user/Desktop/nonlinear.py", "/home/user/Desktop/summary.txt" ], "evaluator": { "postconfig": [ { "type": "upload_file", "parameters": { "files": [ { "local_path": "task-06/eval.py", "path": "/home/user/Desktop/eval.py" } ] } } ], "func": "exact_match", "result": { "type": "vm_command_line", "command": "/home/user/miniconda3/bin/conda run --no-capture-output --prefix /home/user/miniconda3/envs/fenics-legacy /home/user/miniconda3/envs/fenics-legacy/bin/python /home/user/Desktop/eval.py", "shell": "true" }, "expected": { "type": "rule", "rules": { "expected": "True\n" } } }, "proxy": false, "fixed_ip": false, "possibility_of_env_change": "low" }