[yt-users] Slices display incorrectly when many grids are used

Matthew Turk matthewturk at gmail.com
Sun Apr 17 05:46:31 PDT 2011


Hi Nathan,

Are you by any chance placing the slice on a grid boundary?  We
sometimes see this; John ZuHone has seen it with FLASH as well, and it
was solved by jittering the coordinate slightly.  (For what it's worth
FLASH also uses CGS for its coordinates, which I think contributes to
the mismatch.)  For instance,

dx = pf.h.get_smallest_dx()
center = na.zeros(3) + 1e-3 * dx
pc = PlotCollection(pf, center)

I think this comes from the two-step process that identifies cells
that contribute to a slice.  The first step identifies grids that are
intersected by a slice, and the second is cells within those grids.
For some reason in some simulations when you slice right along a grid
boundary, these two give different results.  I think I might know how
to make this process symmetric, but I'm still not sure what the
"right" thing to do is when the slice is inserted right at the grid
boundary.

Let me know if this helps out.

Best,

Matt

On Sat, Apr 16, 2011 at 10:17 PM, Nathan Roth
<nathaniel.roth at berkeley.edu> wrote:
> Hi yt crew,
>
>   I am analyzing some orion data, and I am finding that slices will not
> display correctly for a simulation with 64 coarse AMR grids, although they
> will display just fine if use 8 coarse grids. Outer grids on my slices are
> appearing as blank white, and I've attached an example. I think the data
> itself is fine, and if I plot along a ray in the same slice I see what I
> expect (I've attached that, too). Note that the program units are cgs, which
> is why the density field values on the ray are on the order of 1e-22. I'd be
> happy to provide more details if they are needed.
>
> Thanks,
> Nathan
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>



More information about the yt-users mailing list