Kaitlin and Matt, This will help the eventual support for cylindrical coordinates anyway, since we'll need to know how to open these files. John On Nov 15, 2012, at 5:50 PM, Matthew Turk wrote:

Hi Kaitlin and John,

On Thu, Nov 15, 2012 at 5:49 PM, John ZuHone wrote:

...

Hi Kaitlin,

At the moment yt supports 2D and 3D cartesian data for Athena (the former only very recently in the yt development tree).

True cylindrical coordinate support (for all codes) is planned for yt 3.0. FLASH 2D cylindrical reads in fine, but this is mostly by accident since the data format in the file is the same as 2D cartesian. I'm guessing this means that the Athena vtk files for cylindrical coordinates are set up slightly differently from the cartesian coordinates.

As a quick followup, I *just now* blogged about 3.0 which includes some discussion of non-cartesian coordinates:

http://blog.yt-project.org/post/WhatsUpWith30.html

I think John's idea is a good one, about using the existing code in 2.X as a stepping stone to swapping out the coordinate system in 3.0.

-Matt

...

Are you using the stable or development version of yt? We might be able to get something working along the same lines as FLASH if you have a file you wouldn't mind providing, as a stepping-stone to true support for cylindrical coordinates.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

...

Hi,

I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian?

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

Hi all, I just posted a PR: https://bitbucket.org/yt_analysis/yt/pull-request/343/athena-upgrades that combines some work that Stephanie Tonnesen and I did a while back to fix some various ways of reading in Athena data. It doesn't address the cylindrical data (unless it is somehow does coincidentally!), but I figured there might be a bit of momentum here and I should get this code PR'd before it bitrots away into nothingness. Cheers! Sam On Thu, Nov 15, 2012 at 6:05 PM, Kaitlin Kratter < kaitlin.kratter@jila.colorado.edu> wrote:

Hi John,

I've put the inputs file and the first output file in vtk and binary format at:

http://jila.colorado.edu/~kakr3691/Athena/

I'm using the development branch (though not necessarily up to date). One thing Matt and I discussed was that using the athena binary data might ultimately be better than the vtk files. My uneducated guess is that the binary data would read in like the 2D cartesian, much like Flash.

-Kaitlin

On Nov 15, 2012, at 3:49 PM, John ZuHone wrote:

Hi Kaitlin,

At the moment yt supports 2D and 3D cartesian data for Athena (the former only very recently in the yt development tree).

True cylindrical coordinate support (for all codes) is planned for yt 3.0. FLASH 2D cylindrical reads in fine, but this is mostly by accident since the data format in the file is the same as 2D cartesian. I'm guessing this means that the Athena vtk files for cylindrical coordinates are set up slightly differently from the cartesian coordinates.

Are you using the stable or development version of yt? We might be able to get something working along the same lines as FLASH if you have a file you wouldn't mind providing, as a stepping-stone to true support for cylindrical coordinates.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

Hi,

I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian?

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

Hi John, Nope, I've never done any work with binary Athena files. Sam On Thu, Nov 15, 2012 at 7:03 PM, John ZuHone wrote:

Hi Sam,

I'll definitely be up for testing this.

Kaitlin,

I agree that we should support binary files as well. I'd be happy to work on that (Sam, I'm not sure if you've done any legwork on this already).

Best,

John

On Nov 15, 2012, at 6:14 PM, Sam Skillman wrote:

Hi all,

I just posted a PR: https://bitbucket.org/yt_analysis/yt/pull-request/343/athena-upgradesthat combines some work that Stephanie Tonnesen and I did a while back to fix some various ways of reading in Athena data. It doesn't address the cylindrical data (unless it is somehow does coincidentally!), but I figured there might be a bit of momentum here and I should get this code PR'd before it bitrots away into nothingness.

Cheers! Sam

On Thu, Nov 15, 2012 at 6:05 PM, Kaitlin Kratter < kaitlin.kratter@jila.colorado.edu> wrote:

...

Hi John,

I've put the inputs file and the first output file in vtk and binary format at:

http://jila.colorado.edu/~kakr3691/Athena/

I'm using the development branch (though not necessarily up to date). One thing Matt and I discussed was that using the athena binary data might ultimately be better than the vtk files. My uneducated guess is that the binary data would read in like the 2D cartesian, much like Flash.

-Kaitlin

On Nov 15, 2012, at 3:49 PM, John ZuHone wrote:

Hi Kaitlin,

At the moment yt supports 2D and 3D cartesian data for Athena (the former only very recently in the yt development tree).

True cylindrical coordinate support (for all codes) is planned for yt 3.0. FLASH 2D cylindrical reads in fine, but this is mostly by accident since the data format in the file is the same as 2D cartesian. I'm guessing this means that the Athena vtk files for cylindrical coordinates are set up slightly differently from the cartesian coordinates.

Are you using the stable or development version of yt? We might be able to get something working along the same lines as FLASH if you have a file you wouldn't mind providing, as a stepping-stone to true support for cylindrical coordinates.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

Hi,

I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian?

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

_______________________________________________ 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

Hi John, Sam, I was using Sam's athena branch. Should I switch to the main development branch, get Sam's PR? Kaitlin On Nov 15, 2012, at 7:34 PM, John ZuHone wrote:

Hi Kaitlin,

Are you using the main development version or Sam's Athena-specific branch? I think we're using different yt code, since we're getting different output. I'm using the main development branch. In that branch, if domain_right_edge doesn't equal -domain_left_edge, you have to set it explicitly.

When I just load the file, I get this:

In [4]: pf = load("cyl_disk.0000.vtk") yt : [INFO ] 2012-11-15 21:28:49,259 Please set 'domain_right_edge' in parameters dictionary argument if it is not equal to -domain_left_edge. yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: domain_dimensions = [ -36.57142857 127.99998696 nan] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 21:28:49,262 Parameters: cosmological_simulation = 0.0

As you can see, the INFO that I get about the domain edges is different from yours.

When I set domain_right_edge explicitly, I get this:

In [5]: pf = load("cyl_disk.0000.vtk", parameters={'domain_right_edge':[15.,np.pi/2.,1.0]}) yt : [INFO ] 2012-11-15 21:25:27,282 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_dimensions = [ 256. 128.00000028 inf] yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: domain_right_edge = [ 15. 1.57079633 1. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: cosmological_simulation = 0.0

So the "inf" in the last domain dimension isn't right, but at least it has the r and z dimensions correct.

I suspect that what has happened is that you might be using a revision of Sam's Athena-specific yt branch which probably does not have 2D support built in, since that was put in just a few days ago. If you use the main development branch and update to the latest revision, you should get the same result that I did above.

I have a feeling I know how to fix the "inf" in the last dimension, but I should probably try Sam's updates that he mentioned and seeing what happens then.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

...

Hi,

I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian?

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

Ok, great. And yes, I have 3D, too. If there is a specific type of data that would be useful to you, I'm of course happy to produce a file. Kaitlin On Nov 16, 2012, at 11:11 AM, John ZuHone wrote:

Hi Kaitlin,

Right now, that is the sequence of events that you should try. Please let us know if you see anything odd. Do you have 3D datasets as well?

Best,

John

On Nov 16, 2012, at 1:07 PM, Kaitlin Kratter wrote:

...

Hi John, Sam,

I was using Sam's athena branch. Should I switch to the main development branch, get Sam's PR?

Kaitlin

On Nov 15, 2012, at 7:34 PM, John ZuHone wrote:

...

Hi Kaitlin,

Are you using the main development version or Sam's Athena-specific branch? I think we're using different yt code, since we're getting different output. I'm using the main development branch. In that branch, if domain_right_edge doesn't equal -domain_left_edge, you have to set it explicitly.

When I just load the file, I get this:

In [4]: pf = load("cyl_disk.0000.vtk") yt : [INFO ] 2012-11-15 21:28:49,259 Please set 'domain_right_edge' in parameters dictionary argument if it is not equal to -domain_left_edge. yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: domain_dimensions = [ -36.57142857 127.99998696 nan] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 21:28:49,262 Parameters: cosmological_simulation = 0.0

As you can see, the INFO that I get about the domain edges is different from yours.

When I set domain_right_edge explicitly, I get this:

In [5]: pf = load("cyl_disk.0000.vtk", parameters={'domain_right_edge':[15.,np.pi/2.,1.0]}) yt : [INFO ] 2012-11-15 21:25:27,282 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_dimensions = [ 256. 128.00000028 inf] yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: domain_right_edge = [ 15. 1.57079633 1. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: cosmological_simulation = 0.0

So the "inf" in the last domain dimension isn't right, but at least it has the r and z dimensions correct.

I suspect that what has happened is that you might be using a revision of Sam's Athena-specific yt branch which probably does not have 2D support built in, since that was put in just a few days ago. If you use the main development branch and update to the latest revision, you should get the same result that I did above.

I have a feeling I know how to fix the "inf" in the last dimension, but I should probably try Sam's updates that he mentioned and seeing what happens then.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

...

Hi,

I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian?

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org

I think that sounds fine. I think Kaitlin was working off my fork to begin with, so it's probably good to go. I think there are probably things I could clean up when I get the time, but I unfortunately don't have much of that right now. On Fri, Nov 16, 2012 at 11:29 AM, John ZuHone wrote:

Hi Kaitlin,

Just go ahead and try the code out on your data to see how it goes, specifically on the 3D data maybe try some projections and whatever other tests you'd like to make sure things look sane. On the basis of that I'll probably accept the PR.

Sam, does that sound good to you?

Best,

John

On Nov 16, 2012, at 1:18 PM, Kaitlin Kratter < kaitlin.kratter@jila.colorado.edu> wrote:

...

Ok, great. And yes, I have 3D, too. If there is a specific type of data that would be useful to you, I'm of course happy to produce a file.

Kaitlin

On Nov 16, 2012, at 11:11 AM, John ZuHone wrote:

...

Hi Kaitlin,

Right now, that is the sequence of events that you should try. Please let us know if you see anything odd. Do you have 3D datasets as well?

Best,

John

On Nov 16, 2012, at 1:07 PM, Kaitlin Kratter < kaitlin.kratter@jila.colorado.edu> wrote:

...

Hi John, Sam,

I was using Sam's athena branch. Should I switch to the main development branch, get Sam's PR?

Kaitlin

On Nov 15, 2012, at 7:34 PM, John ZuHone wrote:

...

Hi Kaitlin,

Are you using the main development version or Sam's Athena-specific branch? I think we're using different yt code, since we're getting different output. I'm using the main development branch. In that branch, if domain_right_edge doesn't equal -domain_left_edge, you have to set it explicitly.

When I just load the file, I get this:

In [4]: pf = load("cyl_disk.0000.vtk") yt : [INFO ] 2012-11-15 21:28:49,259 Please set 'domain_right_edge' in parameters dictionary argument if it is not equal to -domain_left_edge. yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: domain_dimensions = [ -36.57142857 127.99998696 nan] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 21:28:49,262 Parameters: cosmological_simulation = 0.0

As you can see, the INFO that I get about the domain edges is different from yours.

When I set domain_right_edge explicitly, I get this:

In [5]: pf = load("cyl_disk.0000.vtk", parameters={'domain_right_edge':[15.,np.pi/2.,1.0]}) yt : [INFO ] 2012-11-15 21:25:27,282 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_dimensions = [ 256. 128.00000028 inf] yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: domain_right_edge = [ 15. 1.57079633 1. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: cosmological_simulation = 0.0

So the "inf" in the last domain dimension isn't right, but at least it has the r and z dimensions correct.

I suspect that what has happened is that you might be using a revision of Sam's Athena-specific yt branch which probably does not have 2D support built in, since that was put in just a few days ago. If you use the main development branch and update to the latest revision, you should get the same result that I did above.

I have a feeling I know how to fix the "inf" in the last dimension, but I should probably try Sam's updates that he mentioned and seeing what happens then.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

...

Hi,

I gather that cylindrical geometries aren't supported. Is it

possible to force yt to read in cylindrical coordinates as though they were cartesian?

...

When I try to load an athena vtk in cylindrical, I get:

In [32]: pf = load("cyl_disk.0001.vtk") yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting

domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case.

...

yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0

This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15.

Thanks, Kaitlin

_______________________________________________ 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

_______________________________________________ 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

Hi Kaitlin, Can you paste the error somewhere for the merged version? Best, John On Nov 16, 2012, at 4:06 PM, Kaitlin Kratter wrote:

I may have created a merged monster, but:

This version: which is Sam's fork I think merged with the recent PR doesn't break the 3d things I've tested, but doesn't let me save slice plots of the 2D cartesian data. ________ The current version of the code is: 914bda39f0e1+13f17b5eade6+ (yt) tip

But:

The current version of the code is:--- 914bda39f0e1 (yt) tip

Doesn't correctly put together 3D mpi data. It does allow me to save slice plots of the 2D cartesian data.

On Nov 16, 2012, at 11:31 AM, Sam Skillman wrote:

...

I think that sounds fine. I think Kaitlin was working off my fork to begin with, so it's probably good to go. I think there are probably things I could clean up when I get the time, but I unfortunately don't have much of that right now.

On Fri, Nov 16, 2012 at 11:29 AM, John ZuHone wrote: Hi Kaitlin,

Just go ahead and try the code out on your data to see how it goes, specifically on the 3D data maybe try some projections and whatever other tests you'd like to make sure things look sane. On the basis of that I'll probably accept the PR.

Sam, does that sound good to you?

Best,

John

On Nov 16, 2012, at 1:18 PM, Kaitlin Kratter wrote:

...

Ok, great. And yes, I have 3D, too. If there is a specific type of data that would be useful to you, I'm of course happy to produce a file.

Kaitlin

On Nov 16, 2012, at 11:11 AM, John ZuHone wrote:

...

Hi Kaitlin,

Right now, that is the sequence of events that you should try. Please let us know if you see anything odd. Do you have 3D datasets as well?

Best,

John

On Nov 16, 2012, at 1:07 PM, Kaitlin Kratter wrote:

...

Hi John, Sam,

I was using Sam's athena branch. Should I switch to the main development branch, get Sam's PR?

Kaitlin

On Nov 15, 2012, at 7:34 PM, John ZuHone wrote:

...

Hi Kaitlin,

Are you using the main development version or Sam's Athena-specific branch? I think we're using different yt code, since we're getting different output. I'm using the main development branch. In that branch, if domain_right_edge doesn't equal -domain_left_edge, you have to set it explicitly.

When I just load the file, I get this:

In [4]: pf = load("cyl_disk.0000.vtk") yt : [INFO ] 2012-11-15 21:28:49,259 Please set 'domain_right_edge' in parameters dictionary argument if it is not equal to -domain_left_edge. yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:28:49,260 Parameters: domain_dimensions = [ -36.57142857 127.99998696 nan] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:28:49,261 Parameters: domain_right_edge = [-1. 1.570796 -0. ] yt : [INFO ] 2012-11-15 21:28:49,262 Parameters: cosmological_simulation = 0.0

As you can see, the INFO that I get about the domain edges is different from yours.

When I set domain_right_edge explicitly, I get this:

In [5]: pf = load("cyl_disk.0000.vtk", parameters={'domain_right_edge':[15.,np.pi/2.,1.0]}) yt : [INFO ] 2012-11-15 21:25:27,282 Parameters: current_time = 0.0 yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_dimensions = [ 256. 128.00000028 inf] yt : [INFO ] 2012-11-15 21:25:27,283 Parameters: domain_left_edge = [ 1. -1.570796 0. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: domain_right_edge = [ 15. 1.57079633 1. ] yt : [INFO ] 2012-11-15 21:25:27,291 Parameters: cosmological_simulation = 0.0

So the "inf" in the last domain dimension isn't right, but at least it has the r and z dimensions correct.

I suspect that what has happened is that you might be using a revision of Sam's Athena-specific yt branch which probably does not have 2D support built in, since that was put in just a few days ago. If you use the main development branch and update to the latest revision, you should get the same result that I did above.

I have a feeling I know how to fix the "inf" in the last dimension, but I should probably try Sam's updates that he mentioned and seeing what happens then.

Best,

John

On Nov 15, 2012, at 5:36 PM, Kaitlin Kratter wrote:

> > > Hi, > > I gather that cylindrical geometries aren't supported. Is it possible to force yt to read in cylindrical coordinates as though they were cartesian? > > > When I try to load an athena vtk in cylindrical, I get: > > In [32]: pf = load("cyl_disk.0001.vtk") > yt : [INFO ] 2012-11-15 15:31:26,443 Temporarily setting domain_right_edge = -domain_left_edge. This will be corrected automatically if it is not the case. > yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: current_time = 0.001080635 > yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_dimensions = [ -37 128 -2147483648] > yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_left_edge = [ 1. -1.570796 0. ] > yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: domain_right_edge = [-1. 1.570796 -0. ] > yt : [INFO ] 2012-11-15 15:31:26,445 Parameters: cosmological_simulation = 0.0 > > > This is a 256 by 128 grid, and the dimensions should be -pi/2, pi/2 and 1, 15. > > > Thanks, > Kaitlin > > > > _______________________________________________ > 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

_______________________________________________ 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

_______________________________________________ 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