Re: FLRNDM question

From: Giuseppe Battistoni (Giuseppe.Battistoni@mi.infn.it)
Date: Wed Apr 11 2007 - 12:36:16 CEST

  • Next message: Rollet Sofia: "flair"

    Dear Frances,
    You are always starting from the same random seed (which is the
    random.dat file inside your $FLUPRO directory).w

    You can obtain a different initial seed in two ways:
    1) Filling a value in the WHAT(2) of the RANDOMIZ card (see the manual and
    the last part of the Note in the description of this command)

    2) You can collect somewhere a certain number of output seeds of FLUKA
    (the files called ran**002 ran**03 etc.).
    These can be used either to rename the random.dat file which is locate in
    $FLUPRO, or (much better) they can be renamed according to your input file
    and called by the rfluka script.
    Let me make an example:
    suppose that you are using an input file called frances.inp
    Notice that you will obtain, as output seed from the run #1, a file
    called ranfraces002.
    At that point if you run again with
    ./flutil/rfluka -N1 -M2 frances
    Then you will use ranfrances002 as a start instead of the default
    random.dat (which was copied unto ranfrances001), provided that you are
    including the RANDOMIZ.
    Alternatively you can rename ranfrances002 into ranfrances001 and run with
    the command:
    ./flutil/rfluka -N0 -M1 frances

    Ciao
            Giuseppe

    On Wed, 11 Apr 2007, Burge, F (Frances) wrote:

    > Dear all,
    >
    > I have been using the FLRNDM random number generator to sample from a
    > spectrum in the source subroutine. To check that everything works as it
    > should, I have been running for a small number of particles (typically
    > 10) and writing the sampled energy etc to a separate file. The energy is
    > sampled according to the following statement:
    >
    > RANDNO(NPFLKA) = FLRNDM(DUM)
    > TKEFLK(NPFLKA) = (EMIN + (EMAX - EMIN) * RANDNO(NPFLKA))
    >
    > I have noticed that every time I run the program, I get the same
    > energies sampled for the ten particles, which is a result of the same
    > random numbers being used each time. Is there something I need to
    > include to make sure that I get different sequences on each run?
    >
    > thanks,
    >
    > Frances
    >
    > Frances Burge
    > Health Physicist
    > 01235 778269
    > Diamond Light Source Ltd.
    > Rutherford Appleton Laboratory
    > Chilton
    > Didcot
    > OX11 0DE
    >
    >


  • Next message: Rollet Sofia: "flair"

    This archive was generated by hypermail 2.1.6 : Wed Apr 11 2007 - 12:48:03 CEST