You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
is it possible to add the true_solution in the stokes problem?
and would it be possible to visualise the residual in 2d? Similarly to poisson?
or else, what visualisation do you suggest for evaluation ?
Best
The text was updated successfully, but these errors were encountered:
Hi @dgm2 ,
It is not possible to state the analytical true solution of the stokes problem but for few cases documented in literature.
You could by the way compare the results arising from PINN solvers with a reference one, obtained via traditional numerical methods (you can easily find some Python code on GitHub for those).
I would then suggest for evaluation to visualize the magnitude of u (namely, the l2 norm of u) for the prediction, for the reference solution, and then the error between the two. Then you can do the same for p.
Hope it can help a bit!
Hello
is it possible to add the true_solution in the stokes problem?
and would it be possible to visualise the residual in 2d? Similarly to poisson?
or else, what visualisation do you suggest for evaluation ?
Best
The text was updated successfully, but these errors were encountered: