Cylindrical and Polar Slicing (FLASH)
Dear yt gurus, I've enjoyed using yt for a number of FLASH problems in cartesian geometry. Right now I'm just starting on a setup in FLASH 2D polar (eventually also in 3D cylindrical) geometries. These are ordered 2D polar: r, theta 3D cylindrical: r, theta, z I was intrigued when I noticed this image on the yt-blog ( http://blog.yt-project.org/post/WhatsUpWith30.html) -- which is almost exactly what I would like to do. [image: http://blog.yt-project.org/attachments/cylindrical_pixelizer.png] I have updated to yt-3.0, and tried the simplest "Sliceplot" of the data: fn = "const_rho_g_atm_hdf5_plt_cnt_0013" pf = load(fn) fig = SlicePlot(pf, 2,'dens').save("myfig.png") This reads loads the 2D polar plot file correctly, but renders the result in a cartesian sense. Does anyone have advice/recipes for how to proceed with Slices rendered "cylindrically"? I understand that this feature set is not yet fully developed but I was curious if there was a solution (or parts of a solution) floating around out there. Thanks so much, Morgan MacLeod ------------------------------------------------- Graduate Student Astronomy & Astrophysics UC Santa Cruz
Hi Morgan, Welcome to yt! On Thu, Aug 15, 2013 at 8:58 PM, Morgan MacLeod <morganmacleod@gmail.com>wrote:
Dear yt gurus,
I've enjoyed using yt for a number of FLASH problems in cartesian geometry. Right now I'm just starting on a setup in FLASH 2D polar (eventually also in 3D cylindrical) geometries.
These are ordered 2D polar: r, theta 3D cylindrical: r, theta, z
I was intrigued when I noticed this image on the yt-blog ( http://blog.yt-project.org/post/WhatsUpWith30.html) -- which is almost exactly what I would like to do.
[image: http://blog.yt-project.org/attachments/cylindrical_pixelizer.png]
I have updated to yt-3.0, and tried the simplest "Sliceplot" of the data:
fn = "const_rho_g_atm_hdf5_plt_cnt_0013" pf = load(fn) fig = SlicePlot(pf, 2,'dens').save("myfig.png")
This reads loads the 2D polar plot file correctly, but renders the result in a cartesian sense.
Does anyone have advice/recipes for how to proceed with Slices rendered "cylindrically"? I understand that this feature set is not yet fully developed but I was curious if there was a solution (or parts of a solution) floating around out there.
Thanks so much,
Morgan MacLeod
So, this is an area of active development. Right now, you can do it *manually* by calling the pixelizer, but unfortunately it's not automated yet -- Anthony, JohnZ and I almost got it completely integrated to seamlessly do this, but unfortunately work stalled for a while. Today and tomorrow I have to attend to some release & conference issues, but because I think this is quite close to completion I will be putting an item on my todo list to attempt to address this on Monday, including providing examples and tests. Best, Matt
------------------------------------------------- Graduate Student Astronomy & Astrophysics UC Santa Cruz
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Morgan, Sorry for the delay in replying. I've spent some time looking at the polar coordinate handling in yt-3.0 this morning and I believe I have come up with the current status. As it stands, the actual creation of a nice plot like yt normally does -- through the PlotWindow object -- is not yet possible. There are a few more steps that need to happen before everything can be nicely integrated like that. However, I have refreshed my memory on the coordinate handling system, and I have been able to make some nice images relatively easily. The code had to have a few changes to support this; I've put these changes into a pull request here: https://bitbucket.org/yt_analysis/yt-3.0/pull-request/82/fix-flash-polar-cyl... You can get them by doing: cd $YT_DEST/src/yt-hg/ (or wherever your yt-3.0 installation is) and doing: hg pull -r 12101b782ac8 https://bitbucket.org/MatthewTurk/yt-3.0 hg up -r 12101b782ac8 python2.7 setup.py build_ext -i -f (the -f will force every extension to rebuild; if you get funny errors with selector objects being the wrong size, you may need to do: "find . -name '*.pyx' | xargs touch" and then another build_ext run.) This script then created the correct output for me: http://paste.yt-project.org/show/3788/ I've tested it on a few NIF items. The changes I made were specifically in how the data is expected to be ordered; I'd incorrectly had cylindrical as rtz and polar as rzt in FLASH. I've switched these in the PR. Roman, Suoqing and Anthony, I've listed you as reviewers on the PR because I know you have all looked at polar and/or cylindrical data in the past. If you could give some feedback, and especially let me know if I have done something wrong. In the future, I would like to have things working such that we can do SlicePlot( ... ) like normal. That will require a somewhat detailed discussion with PlotWindow experts about how to set up bounds for a window, but I think it should be doable in the not-too-distant future. Let me know if that works for you, and if you need any more help! Best, Matt On Thu, Aug 15, 2013 at 8:58 PM, Morgan MacLeod <morganmacleod@gmail.com>wrote:
Dear yt gurus,
I've enjoyed using yt for a number of FLASH problems in cartesian geometry. Right now I'm just starting on a setup in FLASH 2D polar (eventually also in 3D cylindrical) geometries.
These are ordered 2D polar: r, theta 3D cylindrical: r, theta, z
I was intrigued when I noticed this image on the yt-blog ( http://blog.yt-project.org/post/WhatsUpWith30.html) -- which is almost exactly what I would like to do.
[image: http://blog.yt-project.org/attachments/cylindrical_pixelizer.png]
I have updated to yt-3.0, and tried the simplest "Sliceplot" of the data:
fn = "const_rho_g_atm_hdf5_plt_cnt_0013" pf = load(fn) fig = SlicePlot(pf, 2,'dens').save("myfig.png")
This reads loads the 2D polar plot file correctly, but renders the result in a cartesian sense.
Does anyone have advice/recipes for how to proceed with Slices rendered "cylindrically"? I understand that this feature set is not yet fully developed but I was curious if there was a solution (or parts of a solution) floating around out there.
Thanks so much,
Morgan MacLeod ------------------------------------------------- Graduate Student Astronomy & Astrophysics UC Santa Cruz
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hello Matt, Thank you! That works very nicely. I was able to plot a test 2D polar simulation by commenting line 201 of geometry/coordinate_handler.py def pixelize(self, dimension, data_source, field, bounds, size, antialias = True): #raise NotImplimentedError which appears in PolarCoordinateHandler. I really appreciate the help, and the quick reply. Let me know if it is helpful for me to send a small (68k) very low-res 2D polar snapshot for your testing purposes. Morgan On Mon, Aug 19, 2013 at 9:10 AM, Matthew Turk <matthewturk@gmail.com> wrote:
Hi Morgan,
Sorry for the delay in replying. I've spent some time looking at the polar coordinate handling in yt-3.0 this morning and I believe I have come up with the current status.
As it stands, the actual creation of a nice plot like yt normally does -- through the PlotWindow object -- is not yet possible. There are a few more steps that need to happen before everything can be nicely integrated like that. However, I have refreshed my memory on the coordinate handling system, and I have been able to make some nice images relatively easily.
The code had to have a few changes to support this; I've put these changes into a pull request here:
https://bitbucket.org/yt_analysis/yt-3.0/pull-request/82/fix-flash-polar-cyl...
You can get them by doing:
cd $YT_DEST/src/yt-hg/
(or wherever your yt-3.0 installation is) and doing:
hg pull -r 12101b782ac8 https://bitbucket.org/MatthewTurk/yt-3.0 hg up -r 12101b782ac8 python2.7 setup.py build_ext -i -f
(the -f will force every extension to rebuild; if you get funny errors with selector objects being the wrong size, you may need to do: "find . -name '*.pyx' | xargs touch" and then another build_ext run.)
This script then created the correct output for me: http://paste.yt-project.org/show/3788/
I've tested it on a few NIF items. The changes I made were specifically in how the data is expected to be ordered; I'd incorrectly had cylindrical as rtz and polar as rzt in FLASH. I've switched these in the PR.
Roman, Suoqing and Anthony, I've listed you as reviewers on the PR because I know you have all looked at polar and/or cylindrical data in the past. If you could give some feedback, and especially let me know if I have done something wrong.
In the future, I would like to have things working such that we can do SlicePlot( ... ) like normal. That will require a somewhat detailed discussion with PlotWindow experts about how to set up bounds for a window, but I think it should be doable in the not-too-distant future.
Let me know if that works for you, and if you need any more help!
Best,
Matt
On Thu, Aug 15, 2013 at 8:58 PM, Morgan MacLeod <morganmacleod@gmail.com>wrote:
Dear yt gurus,
I've enjoyed using yt for a number of FLASH problems in cartesian geometry. Right now I'm just starting on a setup in FLASH 2D polar (eventually also in 3D cylindrical) geometries.
These are ordered 2D polar: r, theta 3D cylindrical: r, theta, z
I was intrigued when I noticed this image on the yt-blog ( http://blog.yt-project.org/post/WhatsUpWith30.html) -- which is almost exactly what I would like to do.
[image: http://blog.yt-project.org/attachments/cylindrical_pixelizer.png]
I have updated to yt-3.0, and tried the simplest "Sliceplot" of the data:
fn = "const_rho_g_atm_hdf5_plt_cnt_0013" pf = load(fn) fig = SlicePlot(pf, 2,'dens').save("myfig.png")
This reads loads the 2D polar plot file correctly, but renders the result in a cartesian sense.
Does anyone have advice/recipes for how to proceed with Slices rendered "cylindrically"? I understand that this feature set is not yet fully developed but I was curious if there was a solution (or parts of a solution) floating around out there.
Thanks so much,
Morgan MacLeod ------------------------------------------------- Graduate Student Astronomy & Astrophysics UC Santa Cruz
_______________________________________________ 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 Morgan, On Mon, Aug 19, 2013 at 6:59 PM, Morgan MacLeod <morganmacleod@gmail.com>wrote:
Hello Matt,
Thank you! That works very nicely.
Great! Let me know if anything else shows up as an issue. I actually have been working recently on cylindrical coordinates (and on volume rendering cylindrical volumes) so this is an area of interest for me.
I was able to plot a test 2D polar simulation by commenting line 201 of geometry/coordinate_handler.py def pixelize(self, dimension, data_source, field, bounds, size, antialias = True): #raise NotImplimentedError
which appears in PolarCoordinateHandler.
Ah, of course -- I missed that in my pull request. I will update it with this as well.
I really appreciate the help, and the quick reply. Let me know if it is helpful for me to send a small (68k) very low-res 2D polar snapshot for your testing purposes.
Right now I have only "cylindrical" datasets (I think) so this would be *very* helpful. Thank you! Best, Matt
Morgan
On Mon, Aug 19, 2013 at 9:10 AM, Matthew Turk <matthewturk@gmail.com>wrote:
Hi Morgan,
Sorry for the delay in replying. I've spent some time looking at the polar coordinate handling in yt-3.0 this morning and I believe I have come up with the current status.
As it stands, the actual creation of a nice plot like yt normally does -- through the PlotWindow object -- is not yet possible. There are a few more steps that need to happen before everything can be nicely integrated like that. However, I have refreshed my memory on the coordinate handling system, and I have been able to make some nice images relatively easily.
The code had to have a few changes to support this; I've put these changes into a pull request here:
https://bitbucket.org/yt_analysis/yt-3.0/pull-request/82/fix-flash-polar-cyl...
You can get them by doing:
cd $YT_DEST/src/yt-hg/
(or wherever your yt-3.0 installation is) and doing:
hg pull -r 12101b782ac8 https://bitbucket.org/MatthewTurk/yt-3.0 hg up -r 12101b782ac8 python2.7 setup.py build_ext -i -f
(the -f will force every extension to rebuild; if you get funny errors with selector objects being the wrong size, you may need to do: "find . -name '*.pyx' | xargs touch" and then another build_ext run.)
This script then created the correct output for me: http://paste.yt-project.org/show/3788/
I've tested it on a few NIF items. The changes I made were specifically in how the data is expected to be ordered; I'd incorrectly had cylindrical as rtz and polar as rzt in FLASH. I've switched these in the PR.
Roman, Suoqing and Anthony, I've listed you as reviewers on the PR because I know you have all looked at polar and/or cylindrical data in the past. If you could give some feedback, and especially let me know if I have done something wrong.
In the future, I would like to have things working such that we can do SlicePlot( ... ) like normal. That will require a somewhat detailed discussion with PlotWindow experts about how to set up bounds for a window, but I think it should be doable in the not-too-distant future.
Let me know if that works for you, and if you need any more help!
Best,
Matt
On Thu, Aug 15, 2013 at 8:58 PM, Morgan MacLeod <morganmacleod@gmail.com>wrote:
Dear yt gurus,
I've enjoyed using yt for a number of FLASH problems in cartesian geometry. Right now I'm just starting on a setup in FLASH 2D polar (eventually also in 3D cylindrical) geometries.
These are ordered 2D polar: r, theta 3D cylindrical: r, theta, z
I was intrigued when I noticed this image on the yt-blog ( http://blog.yt-project.org/post/WhatsUpWith30.html) -- which is almost exactly what I would like to do.
[image: http://blog.yt-project.org/attachments/cylindrical_pixelizer.png]
I have updated to yt-3.0, and tried the simplest "Sliceplot" of the data:
fn = "const_rho_g_atm_hdf5_plt_cnt_0013" pf = load(fn) fig = SlicePlot(pf, 2,'dens').save("myfig.png")
This reads loads the 2D polar plot file correctly, but renders the result in a cartesian sense.
Does anyone have advice/recipes for how to proceed with Slices rendered "cylindrically"? I understand that this feature set is not yet fully developed but I was curious if there was a solution (or parts of a solution) floating around out there.
Thanks so much,
Morgan MacLeod ------------------------------------------------- Graduate Student Astronomy & Astrophysics UC Santa Cruz
_______________________________________________ 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
participants (2)
-
Matthew Turk -
Morgan MacLeod