[yt-users] h5py execution time

[Jean-Claude Passy]

  Hi all,

I have a unigrid 256^3 simulation that run on 64 processors. I need to 
copy/modify some data: I want to set the velocity of all cells having a 
density higher than a certain threshold to a fixed value, the others to 0.
I wrote a little script using h5py. I have shortened it below:

###################################################
for grid in glob.glob('%s/*.cpu*' %(folderIN+NameIN)):

   file1 = H.File(...,'r')

   file2 = H.File(..., 'w')

   for group in file1.keys():

     mygroup = file1[group]

     subgroup = file2.create_group(group)

*density = file1[group]['Density']*

     for field in mygroup.keys():

       print 'Field:', field, ' in ', mygroup

       mydset = mygroup[field]

       shape = mydset.shape

       where = '/'+group+'/'+field


# Setting up the gas velocity

       if (field == 'x-velocity'):

         tmp = N.zeros(file1[group][field].shape)

         size = file1[group][field].shape

         for k1 in range(0,size[0],1):

           for k2 in range(0,size[1],1):

             for k3 in range(0,size[2],1):

* if (density[k1,k2,k3] > Threshold):*

                 tmp[k1,k2,k3] = SomeValue

               else:

                 tmp[k1,k2,k3] = 0.0

         print 'Setting up Vx = ', SomeValue, 'if rho >', Threshold

         dset = subgroup.create_dataset(field,shape,data=tmp)

###################################################

The issue is that this script needs more than 40 minutes to complete 
!!!! In the meantime, if I replace the test :

################################
*if (density[k1,k2,k3] > Threshold):*
################################

by a static test like:

################################
*if (10 > Threshold):*
################################

then the execution lasts only a few seconds ! So the problem comes from 
the test involving the field density.

I don't really understand what is going on since I declared the variable 
density outside every loop. Could it be a memory issue ?
I would really appreciate if someone could help solve that problem.

Thanks a lot,


JC
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20100812/c3bb1da8/attachment.htm>