On Mon, Feb 4, 2019 at 6:05 AM Aman Saxena via petsc-users < petsc-users@mcs.anl.gov> wrote:
> I am trying to use PETSc's TS to implement implicit time-stepping, for > 1d-Euler Equation discretized using 5th order WENO schemes. I have > following queries related to using some of the snes options:: > > -- What does -ts_max_snes_failure means and what does the non-linear > solver do once it diverges to ensure convergence? > You want to look at the manual but I think this is the number of nonlinear solver failures that it will try to recover from before quitting. > > -- If I do not specify “-snes_fd” or “-snes_mf”, does by default it uses > the jacobian I have coded instead of building it using Finite-differences? > test it but if you give it a Jacobian I would hope that it uses it! > > -- As “-snes_fd ” replaces my routine to calculate the jacobian with > petsc’s fd routine, does it stores the jacobian in the same Matrix that I > pass to my hand-coded routine. It might lead to some inconsistency as my > actual scheme involves stencils width equals 3 while I am approximating the > Jacobian using first order scheme which requires stencils width equals 1. > So I have used another “dmda” with sw = 1, to initialize the matrix using > “DMCreateMatrix()” while fd needs sw = 3. > This should be OK. Try it. > > -- I am not able to use “-snes_type test” it gives me the following error > “[0]PETSC ERROR: Unknown type. Check for miss-spelling or missing package: > http://www.mcs.anl.gov/petsc/documentation/installation.html#external > [0]PETSC ERROR: Unable to find requested SNES type test” > > This has changed. Use -snes_test_jacobian > -- What exactly "PCMAT" do? Do I have to supply matrix other than in > RHSJacobian for PCMAT? > I'm not sure where you picked this up you don't need to worry about it: https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/PC/PCMAT.html > > Thanks >