Working on all vertebrae in parallel

[idhamari]

This should reduce the time to complete spine segmentation but it seems Slicer python doesn't support run in parallel.

I tried this:

   jobs=[0,1,2,3,4,5,6,7]       
   if isParallel ==1:
      results1 = []
      print '--pooling starts now--'
      pool = Pool( processes=8 )
      r1 = pool.map_async(self.runCroppingAll,jobs, callback=results1.append)
      r1.wait() # Wait for the results
      print '--pooling completed--'
      raw_input("Press Enter to continue...")

but either Slice rcrashes or hang.

[idhamari]

next to try:

1. prepare all parameters into txt file.
2. run only one command that run external python script
   • the script read the text file and apply the process in parallel.

tried system python and slicer python in additon to

        Slicer --no-main-window --python-script  file.py

but nothing work. It is mentioned here that PyQT does not support parallel process.

[idhamari]

next to try: try to call a compiled cpp or shell script that call in parallel.

[idhamari]

tried calling shell script but does not work. What work is:

cmds = c1 + " & " + c2 + " & " + c3 + " & " + c4 + " & " + c5 + " & " + c6 + " & " + c7
os.system( cmds)

problem: how to get the same command in windows?

tested in windows and works:

  cc = " & start /b "
  ce = " )| pause"
  cmds ="( start " +  cmd1 + cc + cmd2 + cc + cmd3 + cc + cmd4 + cc + cmd5 + cc + cmd6 + cc + cmd7 + ce

next step, implement this solution and test it then close this issue.