Instructions to Run the Program

• In your usranal.sf, include the lines:

  integer skmevnt

  common /drvskim/ skmevnt

• Still in your usranal.sf, set the variable skmevnt=1 whenever you want the current event to be skimmed into output (the variable skmevnt is automatically reset to 0 by the main program every time a new event is read from the input stream).

• In your e831job.nam namelist file, set the output flag outflag=1 and select the transfer or pack option for all the blocks you wish to have on the output stream.

• In your e831job.run running script, choose an appropriate name and location for the output file (logical name OTAPE ).

• Go!

Writing to multiple output streams is almost as easy as writing to a single stream. Currently we support up to 16 output streams.

In addition to the above instructions, you must:

• In your usranal.sf, include the lines:

  $$include 'ioparam.inc'
  $$include 'usrout.inc'

• Also in your usranal.sf, set the variable nostream to the number of streams you want to write this event to.
  Set the variables stream(1:nostream) to the stream numbers you want to write to. For instance, say you have five streams total and the current event has to go to streams 1, 3, and 5:
    nostream=3
    stream(1)=1
    stream(2)=3
    stream(3)=5
    skmevnt=1
  (the variable nostream is automatically reset to 1 and stream(1) is also automatcally set to 1 by the main program every time a new event is read from the input stream). In other words the program returns to the default mode of operation after each event.

• In your e831job.nam namelist file, set the mxstreams flag mxstream=n where n is the total number of files you will write. mxstream must be 5 or less.

• In your e831job.run running script, choose appropriate names and locations for the output files (logical names OTAPE, OTAPE2, OTAPE3, ..., OTAPE10, OTAPE11, ... ).

In your namelist file e831job.nam:

• Set the number of files you want to read: nfiles=3
• Set the first file you want to read: fstfile=1
• Set the number of events per file : nevents=100
• Set the first event (same for every file): fstevent=10 (the job will then read events 10 through 109, for each file)
• Specify the file names in the following way:

  $FILES

  filelist(1)='/balena/file_01.dat'

  filelist(2)='/balena/file_02.dat'

  filelist(3)='/balena/file_03.dat'

  $END

• setenv ITAPE /home/pinocchio

• You can split the full file name between the ITAPE environmental variable and file array filelist(...) in any way - you can even leave ITAPE completely blank.
• If you just want to read a single file, the easiest way to do it is to specify the full name as ITAPE and leave the array filelist(...) empty.
• Do not use ~ in the filenames!
• If you specify a list of N files to be read, you can start the job at file number n by setting fstfile=n . Note tough that the total number of files to be read must then be nfiles<=N-n , otherwise at the end of the job you will get an error when the program tries to open a file which doesn't exist.
• Currently a tape is considered by the Template as a single big file. This means that:
  • You will not be able to skip files starting the job in the middle of the tape (you can skip event tough using the flag "fstevent")
  • For a tape job, nevents is the total number of events to be read, instead of the number of events per file
• Do not use nfiles=0 when reading disk files! This will cause the program to go over the same files over and over again. Set instead nfiles to the exact number of files you want to read. You can use nfiles=0 when running on a tape.
• Despite what history has taught us, Pinocchio was a nerdish computer gik who had a love story over the internet with a girl named "balena"

• Linking: modify drve831.make as follows:
  • use DLIB instead of RLIB everywhere in the file
  • use DINCS instead of RINCS everywhere in the file
• Running: modify e831job.run as follows:
  • source ~e831code/utilities/e831_setenv.csh dev