DELPHI mass production
File naming rules
- External file names must follow the convention
generator_[anything[_anything]]_energy[_mass[_mass]]_run.dat
and should be stored on a readable CASTOR directory
- Capital letters in the file names are not forbidden, but they will not be propagated into the nicknames.
To avoid confusion due to this, please use only lower case characters.
- the generator identifier should include both the generator used and the version,
Example: hzha03pyth6156
- the second and third argument (anything in the notation above) may be omitted,
if given they MUST NOT BE NUMERIC, and are interpreted as channel
- use the reserved key word gas
for the argument just before energy to indicate that this simulation has been
done using RICH Gas setup. Default is liq. This rule implies that you should not use the
the key word gas as channel identifier ! Gas setup is only supported for 93d and 92e simulation.
- the first numeric part in the file name is interpreted as Center-Of-Mass energy. This must be given.
- two other numeric arguments may follow the energy
- the last part of the name must be the absolute value of the run number
- valid example file names are:
- kk2f414_mumu_204_10001.dat
- hzha03pyth6156_hznu_204_120_22006.dat
- mass productions for 99d and 00d will follow these naming conventions, too
-
Identifier names
The identifier names are determined by the Laboratory, the processing and the file name conventions as given above.
For the above examples the corresponding identfiers created are
- xs_kk2f414mumu_e204_l99_1l_a1
- xs_hzha03pyth6156hznu_e204_m120_l99_1l_a1
-
running production yourself
Users are NOT recommended to run productions themselves. If you have productions which may be usefull for other
people, too, please contact us, and deliver the files to be processed in the format described above. For very
special very small very private productions, you can use runsim .
-
Processing stream for external files or generated input:
script
|
location
|
purpose
|
invokation
|
subextsim
|
/afs/cern.ch/user/d/delmc/MCprod
|
job submission for external file processing
|
living script
|
subsim
|
/afs/cern.ch/user/d/delmc/MCprod
|
job submission for mass production
|
living script
|
mkinputs
|
/afs/cern.ch/user/d/delmc/scripts
|
config file creation for external file processing
|
mkinputs path_to_castordir [outputfile]
|
extprodsim
|
/afs/cern.ch/user/d/delmc/MCprod
|
external file processing using prodsim
|
batch job, using config file
|
prodsim
|
/afs/cern.ch/user/d/delmc/MCprod
|
running generation/delsim/delana/dstana + bookkeeping
|
batch job, help available
|
runsim
|
/afs/cern.ch/group/xx/bin/runsim
|
running delsim/delana/dstana
|
by prodsim
|
-
tools for book keeping and post processing
script
|
location
|
purpose
|
invokation
|
extprodsim_bookkeep.csh
|
/afs/cern.ch/user/d/delmc/scripts
|
missing file identification for external files
|
extprodsim_bookkeep.csh todofile.dat
|
sum2pdl.pl
|
/afs/cern.ch/user/d/delmc/scripts
|
PDL file and html docu production
|
sum2pdl.pl [-h] [-f file.conf] [directory[ directory[ ...]]]
|
checksim
|
/afs/cern.ch/user/d/delmc/scripts/checksim
|
run over a given PDLFILE
|
try "checksim -help" for info
|
DELPHI xshort processing
Some external labs delivere new MC as delana output. To produce extended short dst from it, the following utils
are used. They can also be useful if a dstana reprocessing has to be done from delana output which has been kept.
The code to produce extended short dst is on
/afs/cern.ch/delphi/tasks/short_prod
script
|
location
|
purpose
|
invokation
|
shortify.submit
|
/afs/cern.ch/user/d/delmc/shortprod/shortify.submit
|
job submission
|
living script
|
runxsdst
|
/afs/cern.ch/user/d/delmc/scripts/runxsdst
|
running short production
|
from shortify.submit
|
vid2castor
|
/afs/cern.ch/user/d/delmc/scripts/vid2castor
|
migrate VID to CASTOR
|
living script
|
Online - generation of documentation
The scripts "sum2pdl" is run hourly. It's tasks are:
- STEP I: pdl file creation
-
inspection of specified directories where the production output goes (summary files)
-
creation or updating of pdl input files from the summary file information
-
installation of the created pdl input files in a particular cave directory. From this moment on
new files are accessible by the users via nicknames
-
creation or updating of BatIndex files in each cave. These files summaries the identifiers which
can be found in each cave, and give additional information. The BatIndex files are used to create
documentation.
- STEP II: html file fragments
-
creation of html file fragments. For each nickname such a piece of documentation is produced, in
the offline account. A similar directory structure as for the pdl files is created. The informations
needed to do so are taken from the BatIndex files. Directories without such a file are ignored.
This way of doing things allows to include non-standard nicknames to be included in the documentation
creation. In this case the BatIndex file has to be edited by hand.
-
Inside each directory, a HTMLindex file is created, listing the fragments in the directory, and giving more
information like the number of events. This is done mainly for speed reasons, and allowes efficient searching
for particular samples. This feature is used in the final step, and also the batsearch scripts
uses these informations.
- STEP III: index pages and overview
-
one html documentation file is created summarizing samples matching a given year, processing and category.
Which years, processings and categories are considered is defined in the configuration file of sum2pdl.
-
a global index page is created with links pointing to the files created previously.
Configuration:
The configuration file for sum2pdl can be found on /afs/cern.ch/user/d/delmc/cron/sum2pdl.xml.
With a recent browser which is capable to handle css2 style sheets you can view the information directly
by clicking here.
The tags have the following meaning:
-
Within the
<newcave name="somename"> </newcave>
tags the directories to be inspected are defined.
Each directory should be specified with a
<sumdir dir="somedir" >
tag. The name attribute defines the
cave directory to install the pdl file. BE CAREFUL THAT A PDL FILE DOES NOT ALREADY EXIST IN SOME OTHER
DIRECTORY.
- The
<batcaves> </batcaves>
tags defines the cave directories to be searched for BatIndex files to create
documentation. The special value all matches all existing BatCaves.
- <years>, <versions> and <categories>
define the years,versions and categories to be
included in the global index page.
-
Special pages can be generated, summarizing all files in a particular cave. This is usefull for large scale new
productions. The generated pages are linked from the global index page.
The format is
<bycave cave="somecave"> some comments </bycave>>
-
Additional links to be added to the global index page can be specified with the link tag. The format
is
<link href="some url"> some comments </bycave>>
-
With the info tag additional information can be added for particular nicknames. Within the
<info nickname="xs_my_nickname"> </info>
tags the following tags are allowed:
-
< xsec>some cross section </xsec>
The cross section.
-
<settings href="http://some_url"></settings>
This is meant to link some generator
description pages
-
<target> some text </target>
This was introduced to specify the number of events requested
for running productions. It can be missused to put any comment.
If you have questions or comments send them to delphi-core@cern.ch