Dear Christian, PROOF is under heavy development at the moment and most things of the early prototype will be changed. Don't try using it now. We will come with extensive examples once the thing can be used. To some of your questions. Yes, you will be able to process sets of files and during the processing you can generate new trees that can be analyzed in turn in parallel. More later. Cheers, Fons. On Sun, 2002-03-10 at 19:07, cstrato@EUnet.at wrote: > Dear Rooters > > I am currently trying to understand the PROOF code and feel lost: > It would be really helpful to have some tutorial code. > > A simple(?) and very helpful tutorial would be the following: > First I create a TFile and two TTrees T1 and T2, which I fill > randomly with px and py: gRandom->Rannor(px,py) > These two trees are stored in the TFile. > In a second step I want to read px, py from each tree, and store > the calculated value: pz = px*px + py*py > in two new TTrees T1z and T2z, respectively, which are then also > stored in TFile. > > Using PROOF and/or(?) proofd the creation of trees T1 and T2, > and of T1z and T2z could be done in parallel on two slaves. > > In this respect I have (too) many questions: > How would the code for this macro look like? > > Is it possible to write a program which can create the above > mentioned trees serially on a single processor machine, > but automatically can take advantage of an SMP machine or > two or more machines connected over a network? > > When using the tutorial spyserv.C/spy.C as example: > a, How can spyserv be changed so that two or more slaves > handle the requests of two or more spy clients in parallel? > b, How can spyserv be changed to create independent histograms > on two slaves, displaying them in two different TCanvas? > > Can I test PROOF on my single-processor PowerBook running LinuxPPC? > Is it possible to define only one slave in the config file: > node localhost > slave localhost > > How do I start "proofserv"? > > When do I need to start proofd: proofd -p 1095 $ROOTSYS ? > > Which of the following classes need to be used in user code: > TProof, TProofServ, TProofPlayer, TDSet? > For example: > Do I need to explicitely call TProof::ConnectFile(file)? > Do I need to call TProofServ::ReceiveFile(file)? > Do I need to wrap any object in TDSet for PROOF processing? > When will TDSet be used? > > > Thank you very much in advance. > > Best regards > Christian > ---------------------------------- > C.h.r.i.s.t.i.a.n S.t.r.a.t.o.w.a > V.i.e.n.n.a, A.u.s.t.r.i.a > -- Org: CERN, European Laboratory for Particle Physics. Mail: 1211 Geneve 23, Switzerland E-Mail: Fons.Rademakers@cern.ch Phone: +41 22 7679248 WWW: http://root.cern.ch/~rdm/ Fax: +41 22 7679480
This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:45 MET