Closed epprob closed 1 year ago
Maybe, @clambr when you have done the implementation of the iterative solver, you could post a comparison of the run time of both solvers here (screenshot of profiling, as I did above).
For an iterative solver, I used the PyAMG, a library of Algebraic Multigrid (AMG) solvers. from pyamg import smoothed_aggregation_solver from scipy.sparse.linalg import cg
testcase_inverse_problem.py
for max_nr_of_iterations = 5
Direct Solver
Iterative Solver
Use an iterative or conjugate gradient solver for both blood flow and adjoint equations. Currently, the direct solver requires 96% of run time while solving an inverse problem with 100'000 vertices