Hi all.
This is a projection of the Dark_Matter_Density field from the initial output of a standard enzo cosmological run with just 2 cores. It looks like something is going wrong on the grid boundaries. This occurs for projections and slices, but only for gridded particle data. The gas density looks fine. It seems to occur at every grid boundary, which I tested by increasing the number of cores (so the same number of grids).
I can't tell if this is an enzo or yt issue, but I figured I should ask here first in case I missed something on the yt side. Has anyone seen this before?
Best, Casey
P.S. What is the status of the yt image server? Should we attach or imgur images in the future?
Hi Casey,
On Mon, Feb 27, 2012 at 11:45 PM, Casey W. Stark caseywstark@gmail.com wrote:
Hi all.
This is a projection of the Dark_Matter_Density field from the initial output of a standard enzo cosmological run with just 2 cores. It looks like something is going wrong on the grid boundaries. This occurs for projections and slices, but only for gridded particle data. The gas density looks fine. It seems to occur at every grid boundary, which I tested by increasing the number of cores (so the same number of grids).
This is "real" in the sense that what you're seeing *is* in the data, but in fact is not a problem in the simulation. The grid edges show up because particles *should* but *cannot* be interpolated across these boundaries during yt, and in Enzo to get them to be interpolated at the output time an exchange of boundary conditions has to occur. Adding this to Enzo shouldn't be hard; adding it to yt might be possible if you extend your other patch to work with smoothed covering grids and validate spatial of at least one or two zones.
-Matt
I can't tell if this is an enzo or yt issue, but I figured I should ask here first in case I missed something on the yt side. Has anyone seen this before?
Best, Casey
P.S. What is the status of the yt image server? Should we attach or imgur images in the future?
yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Thanks Matt.
Good to know that it is Enzo and not a bug either. Is there a FAQ on issues like this anywhere? It would be very helpful for new users and I would be happy to document this one.
Best, Casey
On Tuesday, February 28, 2012, Matthew Turk wrote:
Hi Casey,
On Mon, Feb 27, 2012 at 11:45 PM, Casey W. Stark <caseywstark@gmail.comjavascript:;> wrote:
Hi all.
This is a projection of the Dark_Matter_Density field from the initial output of a standard enzo cosmological run with just 2 cores. It looks
like
something is going wrong on the grid boundaries. This occurs for
projections
and slices, but only for gridded particle data. The gas density looks
fine.
It seems to occur at every grid boundary, which I tested by increasing
the
number of cores (so the same number of grids).
This is "real" in the sense that what you're seeing *is* in the data, but in fact is not a problem in the simulation. The grid edges show up because particles *should* but *cannot* be interpolated across these boundaries during yt, and in Enzo to get them to be interpolated at the output time an exchange of boundary conditions has to occur. Adding this to Enzo shouldn't be hard; adding it to yt might be possible if you extend your other patch to work with smoothed covering grids and validate spatial of at least one or two zones.
-Matt
I can't tell if this is an enzo or yt issue, but I figured I should ask
here
first in case I missed something on the yt side. Has anyone seen this before?
Best, Casey
P.S. What is the status of the yt image server? Should we attach or imgur images in the future?
yt-users mailing list yt-users@lists.spacepope.org javascript:; http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
yt-users mailing list yt-users@lists.spacepope.org javascript:; http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Casey,
There's not yet a FAQ on this -- feel free to add it to the yt FAQ or to the Enzo docs under the FAQ section, and it'll geti ncluded in the respective rebuilds.
A few months ago someone claimed to me that they had solved this problem in Enzo with a one or two line modification to the IO routines, by adding a boundary exchange just before output. This fix was never described or passed along. For the purposes of yt I think we can address it more directly by allowing a buffer zone when calculating CIC_Deposit and including particles in the manner similar to the fix you showed in the other thread.
-Matt
On Tue, Feb 28, 2012 at 10:14 AM, Casey W. Stark caseywstark@gmail.com wrote:
Thanks Matt.
Good to know that it is Enzo and not a bug either. Is there a FAQ on issues like this anywhere? It would be very helpful for new users and I would be happy to document this one.
Best, Casey
On Tuesday, February 28, 2012, Matthew Turk wrote:
Hi Casey,
On Mon, Feb 27, 2012 at 11:45 PM, Casey W. Stark caseywstark@gmail.com wrote:
Hi all.
This is a projection of the Dark_Matter_Density field from the initial output of a standard enzo cosmological run with just 2 cores. It looks like something is going wrong on the grid boundaries. This occurs for projections and slices, but only for gridded particle data. The gas density looks fine. It seems to occur at every grid boundary, which I tested by increasing the number of cores (so the same number of grids).
This is "real" in the sense that what you're seeing *is* in the data, but in fact is not a problem in the simulation. The grid edges show up because particles *should* but *cannot* be interpolated across these boundaries during yt, and in Enzo to get them to be interpolated at the output time an exchange of boundary conditions has to occur. Adding this to Enzo shouldn't be hard; adding it to yt might be possible if you extend your other patch to work with smoothed covering grids and validate spatial of at least one or two zones.
-Matt
I can't tell if this is an enzo or yt issue, but I figured I should ask here first in case I missed something on the yt side. Has anyone seen this before?
Best, Casey
P.S. What is the status of the yt image server? Should we attach or imgur images in the future?
yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
participants (2)
-
Casey W. Stark -
Matthew Turk