Hi all, In the ex19, specifically from Lines 581-587, about the "invert the system" part,
581: /* invert the system:582: [ dfu / du 0 0 0 ][yu] = [fu]583: [ 0 dfv / dv 0 0 ][yv] [fv]584: [ dfo / du dfo / dv dfo / do 0 ][yo] [fo]585: [ dft / du dft / dv 0 dft / dt ][yt] [ft]586: by simple back-substitution587: */ I realized u and v are independent variables, but even so, the first two rows should be both containing dfu/dv and dfv/du terms just like the 3rd and 4th rows, do I miss some points? Thank you in advance! On Thu, Oct 2, 2014 at 3:02 PM, Jed Brown <[email protected]> wrote: > Sharp Stone <[email protected]> writes: > > > Hi all, > > > > I'm sorry I'm still confusing that if I have calculated the Jacobian > > elements, how can I go ahead to solve the linear system with dof>1 at > each > > node? The example code ex19 solves the linear system with Guass-Seidel > > method, > > As we just said, the same code (ex19.c) can use many different methods. > The default method is Newton-Krylov with block Jacobi preconditioning. > Run with -snes_view to see details. > -- Best regards, Feng
