[yt-dev] Data selection issue in yt 3.0

[Matthew Turk]

Hi Nathan,

On Mon, Nov 26, 2012 at 10:53 PM, Nathan Goldbaum <goldbaum at ucolick.org> wrote:
> Hi all (specifically Matt, I suspect),
>
> I'm running into an odd issue in yt 3.0.  I'm using the following script:
> http://paste.yt-project.org/show/2905/.  This refers to a non-axissymmetric
> dataset I generated with enzo:
> http://ucolick.org/~goldbaum/files/DD0000.tgz.
>
> The issue is that when I run this script with yt-3.0, I only get back the
> 'z' projection, even in the cases where I ask it for the 'x' and 'y'
> projection.

Yup, it was a typo.  I've fixed it here:

https://bitbucket.org/yt_analysis/yt-3.0/changeset/df11c944851c229698721513fb2bb2f4b6fb61db

Thanks for catching this!  It also reminds me that in the Chunking
YTEP I should mention the fcoords, icoords and fwidth properties of
chunks.

>
> Since the signature of __init__ for the projection object is slightly
> different in 3.0, you'll need to manually choose the line in the script that
> creates the projection object depending on which version of yt you're
> running.  As a side note, is there a reason for this change in the API?

The reason was based on outstanding issues in the way that projections
were different from everything else.  Here's the ticket, filed about a
year and a half ago, where this got talked about:

https://bitbucket.org/yt_analysis/yt/issue/292/consistent-projection-arg-order

At this time I'm no longer certain that changing is the right thing,
since I believe part of the original motivation was for plot
collections to match the projection API.

Thanks for catching the bug; while we do run tests on yt 3.0, it seems
this slipped through.

-Matt

>
> Nathan Goldbaum
> Graduate Student
> Astronomy & Astrophysics, UCSC
> goldbaum at ucolick.org
> http://www.ucolick.org/~goldbaum
>
> _______________________________________________
> yt-dev mailing list
> yt-dev at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
>