[Paraview] Parallel reading of data using pvbatch + python

[Ganesh Vijayakumar]

Hello,

  I'm working with a dataset in tecplot format. Each processor generates a
separate file for its own part of the domain. I read all of them in, merge
the datasets using the group datasets filter and then visualize them (make
contours etc.).

nProcs = len(glob.glob('scalar-*dat'))
allOutData = []

for i in range(1,nProcs+1):
    # create a new 'Tecplot Reader'

allOutData.append(TecplotReader(FileNames=['out-{}-{:05d}.dat'.format(sys.argv[1],i)]))
    allOutData[-1].DataArrayStatus = ['u', 'v', 'w', 'P', 'phi', 'node']

# # create a new 'Group Datasets'
groupDatasets1 = GroupDatasets(Input=allOutData)

How can I speed this up? The reading of the data seems to be the most time
consuming part. So running paraview in parallel is not really helping as
now each processor has to read all the files as opposed to just one
processor eariler.

ganesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20160218/585438d1/attachment.html>