Dear YT users, I am running enzo test problem with MPI. I am unable to visualize the data with YT and it produces following error. If i do not run enzo with MPI (MPICH2) then it works fine. Any ideas, what is going wrong here. Thanks in advance Cheers Latif t : [INFO ] 2012-04-05 17:42:42,770 Parameters: current_time = 7 yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_dimensions = [16 16 16] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_left_edge = [ 0. 0. 0.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_right_edge = [ 1. 1. 1.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: cosmological_simulation = 0.0 yt : [INFO ] 2012-04-05 17:42:42,775 Getting the binary hierarchy Traceback (most recent call last): File "./enzo.py", line 7, in <module> pc=PlotCollection(pf) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/visualization/plot_collection.py", line 108, in __init__ v,self.c = pf.h.find_max("Density") # @todo: ensure no caching File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/static_output.py", line 172, in hierarchy self, data_style=self.data_style) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 217, in __init__ AMRHierarchy.__init__(self, pf, data_style) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/hierarchy.py", line 71, in __init__ self._parse_hierarchy() File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 286, in _parse_hierarchy if self._parse_binary_hierarchy(): return File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 357, in _parse_binary_hierarchy self.grid_dimensions[:] = f["/ActiveDimensions"][:] ValueError: operands could not be broadcast together with shapes (32,3) (8,3)
Hi Latif,
yt stores hierarchy information in between runs. I think what you are
describing is that you have run the same simulation multiple times in
the same location, but you've turned MPI on and off -- this replaces a
portion of the files, but not all. (For this reason, it's recommended
to remove existing output files when you re-run a simulation in Enzo.)
I've put in a change to yt that invalidates the current format in
favor of one that includes a unique hash. This affects all yt users
and will add a small overhead at startup for large hierarchies the
next time they are loaded.
Executing the command "yt update" if you are on the development branch
will update you; if you are not on the development branch, remove the
.harrays files from your directories before loading a new dataset that
has replaced an old one.
Best,
Matt
On Fri, Apr 6, 2012 at 5:21 AM, Abul Latif
Dear YT users, I am running enzo test problem with MPI. I am unable to visualize the data with YT and it produces following error. If i do not run enzo with MPI (MPICH2) then it works fine. Any ideas, what is going wrong here. Thanks in advance Cheers Latif
t : [INFO ] 2012-04-05 17:42:42,770 Parameters: current_time = 7 yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_dimensions = [16 16 16] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_left_edge = [ 0. 0. 0.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_right_edge = [ 1. 1. 1.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: cosmological_simulation = 0.0 yt : [INFO ] 2012-04-05 17:42:42,775 Getting the binary hierarchy Traceback (most recent call last): File "./enzo.py", line 7, in <module> pc=PlotCollection(pf) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/visualization/plot_collection.py", line 108, in __init__ v,self.c = pf.h.find_max("Density") # @todo: ensure no caching File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/static_output.py", line 172, in hierarchy self, data_style=self.data_style) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 217, in __init__ AMRHierarchy.__init__(self, pf, data_style) File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/hierarchy.py", line 71, in __init__ self._parse_hierarchy() File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 286, in _parse_hierarchy if self._parse_binary_hierarchy(): return File "/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py", line 357, in _parse_binary_hierarchy self.grid_dimensions[:] = f["/ActiveDimensions"][:] ValueError: operands could not be broadcast together with shapes (32,3) (8,3)
_______________________________________________ yt-users mailing list yt-users@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
Hi Matt,
Many thanks for quick response!
Indeed, i was running enzo simulations in same directory twice with and
without MPI. I am not on development branch but would adapt the second
approach.
Cheers
Latif
On Fri, Apr 6, 2012 at 11:44 AM, Matthew Turk
Hi Latif,
yt stores hierarchy information in between runs. I think what you are describing is that you have run the same simulation multiple times in the same location, but you've turned MPI on and off -- this replaces a portion of the files, but not all. (For this reason, it's recommended to remove existing output files when you re-run a simulation in Enzo.) I've put in a change to yt that invalidates the current format in favor of one that includes a unique hash. This affects all yt users and will add a small overhead at startup for large hierarchies the next time they are loaded.
Executing the command "yt update" if you are on the development branch will update you; if you are not on the development branch, remove the .harrays files from your directories before loading a new dataset that has replaced an old one.
Best,
Matt
On Fri, Apr 6, 2012 at 5:21 AM, Abul Latif
wrote: Dear YT users, I am running enzo test problem with MPI. I am unable to visualize the data with YT and it produces following error. If i do not run enzo with MPI (MPICH2) then it works fine. Any ideas, what is going wrong here. Thanks in advance Cheers Latif
t : [INFO ] 2012-04-05 17:42:42,770 Parameters: current_time = 7 yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_dimensions = [16 16 16] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_left_edge = [ 0. 0. 0.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: domain_right_edge = [ 1. 1. 1.] yt : [INFO ] 2012-04-05 17:42:42,771 Parameters: cosmological_simulation = 0.0 yt : [INFO ] 2012-04-05 17:42:42,775 Getting the binary hierarchy Traceback (most recent call last): File "./enzo.py", line 7, in <module> pc=PlotCollection(pf) File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/visualization/plot_collection.py",
line 108, in __init__ v,self.c = pf.h.find_max("Density") # @todo: ensure no caching File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/static_output.py",
line 172, in hierarchy self, data_style=self.data_style) File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py",
line 217, in __init__ AMRHierarchy.__init__(self, pf, data_style) File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/data_objects/hierarchy.py",
line 71, in __init__ self._parse_hierarchy() File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py",
line 286, in _parse_hierarchy if self._parse_binary_hierarchy(): return File
"/home/astro73/mlatif/YT/yt-x86_64/src/yt-hg/yt/frontends/enzo/data_structures.py",
line 357, in _parse_binary_hierarchy self.grid_dimensions[:] = f["/ActiveDimensions"][:] ValueError: operands could not be broadcast together with shapes (32,3) (8,3)
_______________________________________________ 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
-- Dr. M. Abdul Latif University of Groningen The Netherlands
participants (2)
-
Abul Latif
-
Matthew Turk