[yt-users] Finding the max density from a checkpoint dump?

[Matthew Turk]

Hi Joseph,

On Tue, May 28, 2013 at 11:07 PM, Joseph Smidt <josephsmidt at gmail.com> wrote:
> Nathan,
>
>   Thanks for your response. I believe I have the lastest version of yt.  I
> get this if I do an hg tip:
>
> [jsmidt at mu-fe1 yt]$ hg tip
> changeset:   8217:fb6cce6a2936
> branch:      yt
> tag:         tip
> user:        Nathan Goldbaum <goldbaum at ucolick.org>
> date:        Mon May 27 09:24:17 2013 -0700
> files:       yt/config.py
> description:
> Bumping the gold standard revision to gold0008.
>
>     Now the checkpoint output is slightly different than a normal data dump.
> for instance, the DD0047/output_0047 file has this comment at the top which
> cannot be read into yt without being removed:
>
> # WARNING! This is a checkpoint dump! Lots of data!
> InitialCycleNumber  = 228
> InitialTime         = 20.046508365013
> InitialCPUTime      = 0
>
>    And there are additional fields like Acceleration fields with these
> checkpoint dumps.  So I was wondering if this other error when finding the
> max values of fields is also checkpoint dump specific as the structure seems
> slightly different compared to normal data dumps.  Thanks again.

I wrote the checkpoint dump mechanism in Enzo, and it wasn't meant to
be used for analysis.  It stores the fluxes (where they exist), the
acceleration, ghost zones, "old" baryon fields, and probably some
other things I've forgotten now.  The idea was to dump all of the
internal data structures that accumulate over the course of a
timestep, and then load them in at a later time, hopefully preserving
the exact state.  (For what it's worth, it never was as successful as
reproducing the exact state of the computation as I had hoped it would
be.)

So my guess is that you're seeing issues with the ghost zones changing
the dimensions of the grids in funny ways.  In many ways, the
checkpoint dumps are more similar to the analysis conducted inline in
Enzo than they are to analysis conducted of standard Enzo data dumps
-- the ghost zones will need to be sliced off and the old/new fields
will need to be interpolated to the centered time.

So to make a long story short, when I use checkpoint dumps in Enzo to
restart the computation, I don't use them for analysis -- I also have
Enzo output standard data dumps and use those.  If we were to try to
make this work, it would need a new IO handler similar to the Inline
IO handler which interpolates between times and slices off ghost
zones.

-Matt

>
> On Tue, May 28, 2013 at 7:45 PM, Joseph Smidt <josephsmidt at gmail.com> wrote:
>>
>> Hey everyone,
>>
>>     My Enzo simulation has refined to enough levels that I need to make
>> checkpoint dumps mid-timestep and do YT analysis from there.  I can read in
>> the checkpoint dump but when I go to find the max density I get this error:
>> http://pastebin.com/2mNVxNyW
>>
>>    Anybody have any idea what is going on and how to fix this?  Also,
>> since this dump is in the middle of the W-cycle, I believe the time on the
>> finest grid is different then the root grid.  How do I find the time on the
>> finest grid for this checkpoint dump?  Thanks.
>>
>> --
>> ------------------------------------------------------------------------
>> Joseph Smidt <josephsmidt at gmail.com>
>>
>> Theoretical Division
>> P.O. Box 1663, Mail Stop B283
>> Los Alamos, NM 87545
>> Office: 505-665-9752
>> Fax:    505-667-1931
>
>
>
>
> --
> ------------------------------------------------------------------------
> Joseph Smidt <josephsmidt at gmail.com>
>
> Theoretical Division
> P.O. Box 1663, Mail Stop B283
> Los Alamos, NM 87545
> Office: 505-665-9752
> Fax:    505-667-1931
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>