[yt-users] Support for boxlib-style orion files

[Matthew Turk]

On Tue, Oct 2, 2012 at 6:56 PM, Mark Krumholz <mkrumhol at ucsc.edu> wrote:
>
> On Oct 2, 2012, at 3:38 PM, Andrew Myers wrote:
>
>> Hi Mark,
>>
>> Exactly what version of yt are you using? I'm on the "tip" of the main development branch (yt-2.x, not the yt-3.0 fork) and I can still load ORION output files, provided the "inputs" file is in my wd. If you're using yt-3.0, I can install that and see if I can reproduce your error.
>>
>> -Andrew Myers
>
> Ahhh, that's the problem! I forgot to put the inputs file in the same directory. It would be nice if the failure message in this case were a little more transparent. It would be even nicer if one didn't actually need the inputs file, since the plot files themselves actually contain all the necessary information, but I guess I can't complain unless I'm willing to implement that capability. ;-)

Hmm, I definitely agree with your comment about the failure message!
As for the inputs file, I think we'd all love to do away with needing
them, and I don't think you need worry about complaining.  :)  I'd
definitely be happy to help or implement the elimination of the inputs
file; I think one reason it was read in initially was to get the top
grid dimensions, which I think can be tricky to extract from the
datafiles without doing a full read.  To my reading of the source, the
parameters that definitely get read from the inputs file are:

amr.n_cell
materials.gamma
amr.ref_ratio
geometry.prob_lo
geometry.prob_hi

Castro/Nyx has a few extras.  In principle we should also be reading:

amr.ref_ratio
geometry.is_periodic

There may be other radiation or FAB parameters, too, that should be read.

-Matt

>
> --
> Mark Krumholz
>
>
>
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org