I use root to read in a large number of particle physics events, apply a load of cuts and then draw some histograms. I do this using the MakeClass method and then putting my cuts inside the Loop() method. I prefer to complie my program rather than run things in cint.
I work on a pc with 4 processors and would like to 1/4 my running time by sharing the load. To start with I would be happy to get 2 processors going. How do I go about doing this? I have looked at PROOF documentation but figure it out.
Can anyone tell me how to do a MakeClass sytle Loop() over all my events but using multiple processors?
See gks05.fzk.de/upload/Exercises/exercises.html for example.
Essentially the only difference with using a TSelector (see the User’s Guide) is to call the proper TProof::Connect.