[Users] init_3_timelevels and nans on past timelevels

[Erik Schnetter]

Most people use either "init_each_timelevel", which calls the initial
data routines several times (with different values in cctk_time), or
"init_fill_timelevels", which initializes one time levels and copies
these data to the past time levels. For TwoPunctions, only the latter
makes sense. This leads to a zeroth order error on the past time
levels, which translates to a first order error on the refinement
boundaries, which translates to a second order error overall. I assume
most people will just accept this error, since initial conditions are
often inaccurate anyway ("junk radiation").

-erik

On Wed, Aug 4, 2021 at 2:21 PM Yosef Zlochower <yosef at astro.rit.edu> wrote:
>
> Thanks Erik,
>
>   What algorithm do you suggest to initialize the past timelevels?
>
>
> On 8/4/21 12:04 PM, Erik Schnetter wrote:
> > CAUTION: This message came from outside RIT. If you are unsure about the source or content of this message, please contact the RIT Service Center at 585-475-5000 or help.rit.edu before clicking links, opening attachments or responding.
> >
> >
> > Yosef
> >
> > As you know, init_3_timelevels uses a special algorithm developed by
> > Scott Hawley (see <https://arxiv.org/abs/gr-qc/0310042>) to initialise
> > 3 time levels with initial conditions when you only have initial
> > conditions for a single time. It uses a carefully orchestrated dance
> > to step forward and backward in time, without having to do time
> > interpolation (which isn't possible yet since there are no other time
> > levels to interpolate from). The end result are (supposed to be) 3
> > time levels full of good initial data that are second order accurate.
> > To my knowledge, this algorithm is rarely used, and over time, it
> > might be that a bug has crept in that would leave certain grid points
> > uninitialized.
> >
> > As you do, I would have recommended using VisIt to analyze the
> > resulting initial conditions and look for nans. I don't know how
> > multiple time levels are mapped to VisIt variables; what you do looks
> > correct. Another option would be to run the NaNChecker with its option
> > to output all nan locations that it finds.
> >
> > -erik
> >
> >
> > On Tue, Aug 3, 2021 at 9:41 PM Yosef Zlochower <yosef at astro.rit.edu> wrote:
> >> Hi,
> >>
> >>     I have been running into an issue where I when I interpolate an
> >> evolved function I get nans even though there are no nans on the
> >> gridfunction itself. The nans appear to arise when the past timelevels
> >> (at iteration 0) are initialized by init_3_timelevels and occur near the
> >> buffer regions. As the evolution proceeds, these bad past timelevels are
> >> overwritten with seemingly good data. Is this the expected behavior of
> >> init_3_timelevels?
> >>
> >> Also, I have been trying to use visit to see what is going on. I used
> >> CarpetIOHDF5::output_all_timelevels = true, and I think this means that
> >> the past timelevels show up as different components of the same
> >> gridfunction although I haven't been able to determine how the component
> >> number is related to the actual components of the grid and the different
> >> timelevels.
> >>
> >> Thanks
> >>
> >>
> >> --
> >> Dr. Yosef Zlochower
> >> Center for Computational Relativity and Gravitation
> >> Associate Professor
> >> School of Mathematical Sciences
> >> Rochester Institute of Technology
> >> 85 Lomb Memorial Drive
> >> Rochester, NY 14623
> >>
> >> Office:74-2067
> >> Phone: +1 585-475-6103
> >>
> >> yosef at astro.rit.edu
> >>
> >> CONFIDENTIALITY NOTE: The information transmitted, including
> >> attachments, is intended only for the person(s) or entity to which it
> >> is addressed and may contain confidential and/or privileged material.
> >> Any review, retransmission, dissemination or other use of, or taking
> >> of any action in reliance upon this information by persons or entities
> >> other than the intended recipient is prohibited. If you received this
> >> in error, please contact the sender and destroy any copies of this
> >> information.
> >>
> >> _______________________________________________
> >> Users mailing list
> >> Users at einsteintoolkit.org
> >> http://lists.einsteintoolkit.org/mailman/listinfo/users
> >
> >
> --
> Dr. Yosef Zlochower
> Center for Computational Relativity and Gravitation
> Associate Professor
> School of Mathematical Sciences
> Rochester Institute of Technology
> 85 Lomb Memorial Drive
> Rochester, NY 14623
>
> Office:74-2067
> Phone: +1 585-475-6103
>
> yosef at astro.rit.edu
>
> CONFIDENTIALITY NOTE: The information transmitted, including
> attachments, is intended only for the person(s) or entity to which it
> is addressed and may contain confidential and/or privileged material.
> Any review, retransmission, dissemination or other use of, or taking
> of any action in reliance upon this information by persons or entities
> other than the intended recipient is prohibited. If you received this
> in error, please contact the sender and destroy any copies of this
> information.
>


-- 
Erik Schnetter <schnetter at cct.lsu.edu>
http://www.perimeterinstitute.ca/personal/eschnetter/