Hi Everyone, I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible? Thanks in advance for your help! Best, Jonah MIller
Hi Jonah, Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual? -Matt On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller < jonah.maxwell.miller@gmail.com> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Matt, Thanks for the reply. The datasets are a six patches coordinate system. I've attached some images of the six data sets, and how they should stitch together. From that, it would be nice to be able to do analysis on them---i.e., make an arbitrary slice plot. It's straightforward to stitch /four of/ the actual arrays together and leave out the top and bottom patches. The problem is getting all six to work together. Best, Jonah On 15-08-20 09:21 AM, Matthew Turk wrote:
Hi Jonah,
Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual?
-Matt
On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller <jonah.maxwell.miller@gmail.com <mailto:jonah.maxwell.miller@gmail.com>> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <mailto: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 Jonah, I've been thinking about this, and I wonder if it's possible to identify which sections overlap. Is there partial overlap between two given cells, or do cells either overlap or *not* overlap? If there's only full overlap in cells (i.e., a cell in one section totally overlaps with a cell in a different section) then I think we can do dynamic or computed masking and create a single unified dataset. Does that make sense? -Matt On Thursday, August 20, 2015, Jonah Miller <jonah.maxwell.miller@gmail.com> wrote:
Hi Matt,
Thanks for the reply. The datasets are a six patches coordinate system. I've attached some images of the six data sets, and how they should stitch together. From that, it would be nice to be able to do analysis on them---i.e., make an arbitrary slice plot.
It's straightforward to stitch *four of* the actual arrays together and leave out the top and bottom patches. The problem is getting all six to work together.
Best, Jonah
On 15-08-20 09:21 AM, Matthew Turk wrote:
Hi Jonah,
Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual?
-Matt
On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller < jonah.maxwell.miller@gmail.com <javascript:_e(%7B%7D,'cvml','jonah.maxwell.miller@gmail.com');>> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing listyt-users@lists.spacepope.org <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');>http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Matt, The data can certainly be arranged so that there's either full overlap or no overlap between the datasets. But I'm not sure what you mean by dynamic/computed masking. Best, Jonah On 15-08-21 10:59 PM, Matthew Turk wrote:
Hi Jonah,
I've been thinking about this, and I wonder if it's possible to identify which sections overlap. Is there partial overlap between two given cells, or do cells either overlap or *not* overlap? If there's only full overlap in cells (i.e., a cell in one section totally overlaps with a cell in a different section) then I think we can do dynamic or computed masking and create a single unified dataset. Does that make sense?
-Matt
On Thursday, August 20, 2015, Jonah Miller <jonah.maxwell.miller@gmail.com <mailto:jonah.maxwell.miller@gmail.com>> wrote:
Hi Matt,
Thanks for the reply. The datasets are a six patches coordinate system. I've attached some images of the six data sets, and how they should stitch together. From that, it would be nice to be able to do analysis on them---i.e., make an arbitrary slice plot.
It's straightforward to stitch /four of/ the actual arrays together and leave out the top and bottom patches. The problem is getting all six to work together.
Best, Jonah
On 15-08-20 09:21 AM, Matthew Turk wrote:
Hi Jonah,
Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual?
-Matt
On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller <jonah.maxwell.miller@gmail.com <javascript:_e(%7B%7D,'cvml','jonah.maxwell.miller@gmail.com');>> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <javascript:_e(%7B%7D,'cvml','yt-users@lists.spacepope.org');> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <javascript:_e(%7B%7D,'cvml','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 Jonah, Sorry, I should have been more explicit. If we regard the dataset as "AMR" where the "AMR" is all on the same level, cells should be able to be overlapped. Can you provide an example set of connectivity, even with fake values, that could be experimented with? On Fri, Aug 21, 2015 at 10:47 PM, Jonah Miller < jonah.maxwell.miller@gmail.com> wrote:
Hi Matt,
The data can certainly be arranged so that there's either full overlap or no overlap between the datasets. But I'm not sure what you mean by dynamic/computed masking.
Best, Jonah
On 15-08-21 10:59 PM, Matthew Turk wrote:
Hi Jonah,
I've been thinking about this, and I wonder if it's possible to identify which sections overlap. Is there partial overlap between two given cells, or do cells either overlap or *not* overlap? If there's only full overlap in cells (i.e., a cell in one section totally overlaps with a cell in a different section) then I think we can do dynamic or computed masking and create a single unified dataset. Does that make sense?
-Matt
On Thursday, August 20, 2015, Jonah Miller <jonah.maxwell.miller@gmail.com> wrote:
Hi Matt,
Thanks for the reply. The datasets are a six patches coordinate system. I've attached some images of the six data sets, and how they should stitch together. From that, it would be nice to be able to do analysis on them---i.e., make an arbitrary slice plot.
It's straightforward to stitch *four of* the actual arrays together and leave out the top and bottom patches. The problem is getting all six to work together.
Best, Jonah
On 15-08-20 09:21 AM, Matthew Turk wrote:
Hi Jonah,
Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual?
-Matt
On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller < jonah.maxwell.miller@gmail.com> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing listyt-users@lists.spacepope.orghttp://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing listyt-users@lists.spacepope.orghttp://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 Matt, Yes, I've uploaded a tarball with a dataset and my preliminary attempts to stitch it together (in an ipython notebook). It also includes a graph showing the connectivity. Here's the file: http://use.yt/upload/d4218c8d Thanks for your help! Best, Jonah On 15-08-30 04:11 PM, Matthew Turk wrote:
Hi Jonah,
Sorry, I should have been more explicit. If we regard the dataset as "AMR" where the "AMR" is all on the same level, cells should be able to be overlapped. Can you provide an example set of connectivity, even with fake values, that could be experimented with?
On Fri, Aug 21, 2015 at 10:47 PM, Jonah Miller <jonah.maxwell.miller@gmail.com <mailto:jonah.maxwell.miller@gmail.com>> wrote:
Hi Matt,
The data can certainly be arranged so that there's either full overlap or no overlap between the datasets. But I'm not sure what you mean by dynamic/computed masking.
Best, Jonah
On 15-08-21 10:59 PM, Matthew Turk wrote:
Hi Jonah,
I've been thinking about this, and I wonder if it's possible to identify which sections overlap. Is there partial overlap between two given cells, or do cells either overlap or *not* overlap? If there's only full overlap in cells (i.e., a cell in one section totally overlaps with a cell in a different section) then I think we can do dynamic or computed masking and create a single unified dataset. Does that make sense?
-Matt
On Thursday, August 20, 2015, Jonah Miller <jonah.maxwell.miller@gmail.com <mailto:jonah.maxwell.miller@gmail.com>> wrote:
Hi Matt,
Thanks for the reply. The datasets are a six patches coordinate system. I've attached some images of the six data sets, and how they should stitch together. From that, it would be nice to be able to do analysis on them---i.e., make an arbitrary slice plot.
It's straightforward to stitch /four of/ the actual arrays together and leave out the top and bottom patches. The problem is getting all six to work together.
Best, Jonah
On 15-08-20 09:21 AM, Matthew Turk wrote:
Hi Jonah,
Right now, I think this would be tricky. I'm trying to figure out precisely how it could be done without compositing the datasets themselves, and I'm not sure it's terribly feasible at the time being without some trickery. One possibility, since the data is spherical, is to get fixed res buffers for each section of the plot you want, then utilize matplotlib to stitch those together into a single plot. It might help if you had a little sketch so that your desired outcome could be a bit more visual?
-Matt
On Wed, Aug 19, 2015 at 2:31 PM, Jonah Miller <jonah.maxwell.miller@gmail.com> wrote:
Hi Everyone,
I have data for a simulation in spherical coordinates that I wish to input into yt and visualize using the generic reader tools. However, the simulation is broken up into six volumes, each of which is a solid angle that makes up part of a sphere. Unfortunately, stitching together the arrays of data produces a lot of redundancies, there's no easy way to include all of it in a single array without including the same data points several times. So what I'd like is a way to feed in each solid angle as an individual data set, but visualize all six datasets on a single plot. Is this possible?
Thanks in advance for your help!
Best, Jonah MIller _______________________________________________ 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 <mailto:yt-users@lists.spacepope.org> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org <mailto: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)
-
Jonah Miller -
Matthew Turk