[Yt-dev] Grid Data Format

Matthew Turk matthewturk at gmail.com
Thu Sep 8 12:01:16 PDT 2011


Awesome work, Sam.

Chris Moody and Casey Stark have both shown an interest in the Grid
Data Format.  I think it would be very useful if we could start
iterating on the format, figuring out where it has weak points, and
see if we can shore it up.  At that point it may be appropriate to
start shopping it around outside this list, as well.  I've invited a
few parties that may be interested in such a discussion to join the
list.

In particular, the GDF will play very nicely with the
deliberate_fields branch, where fields are more clearly defined and
accessed with respect to units, derived fields, and what is in the
data.  Additionally, because GDF separates out particles by type, it
is perhaps a good idea to use this as a starting place to discuss how
we should access particles...  I've asked around a little bit, and
come up with four options:

Option 1: data_object["dark_matter","position_x"]
Option 2: data_object.particles["dark_matter"]["position_x"]
Option 3: data_object.particles["dark_matter","position_x"]
Option 4: data_object["dark_matter"]["position_x"]
(implicit Option 5: ...something else?)

and the response I got was definitely #2 as the favorite.  What do
people here think?

Thanks again, Sam -- this is great.  (And thanks to Chris Moody for
the initial pass at the GDF reader!)

-Matt

On Wed, Sep 7, 2011 at 9:37 PM, Sam Skillman <samskillman at gmail.com> wrote:
> Hi All,
> Today I did some work on the Grid Data Format aka gdf, described here:
> http://yt-project.org/gdf.txt
> I've been attempting to use this to convert an Athena dataset to gdf, then
> read that into yt.  As of right now, I am testing on a single uniform grid,
> and am able to load it, find the max of the fields, and do a few other
> things.  However, I can't seem to get slices/projections to work.  Anyways,
> I'm not sure how much time I'll have in the immediate short term to work on
> this, so if anyone wants to jump in that'd be awesome.
> I've pushed what I have to my fork of yt:
> https://samskillman@bitbucket.org/samskillman/yt
> Here is the test dataset:
> http://corvette.colorado.edu/~skillman/dload/gdf0001.h5
> You should be able to do:
> yt load gdf0001.h5
> As a warning, I haven't tested a bunch of stuff, so there may be stupid
> mistakes lurking around.  Feel free to email me or respond to this list if
> you have questions.
> Cheers,
> Sam
>
> _______________________________________________
> Yt-dev mailing list
> Yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>
>



More information about the yt-dev mailing list