d2wired - DELPHI DST to WIRED Conversion

Installation and Use

/afs/cern.ch/user/a/adye/public/d2wired/

Here are some brief intructions on building and running d2wired. I use HPPLUS, though the program should work on other platforms (perhaps modulo some changes to paths in the scripts). If you have any problems, questions, or suggestions (I already have plenty of these!), please feel free to contact me.

1. Copy the files to your own directory (don't worry about the d2wired subdirectories). If you make any modifications, please let me know - there are several people interested in this program so I'd like to keep all improvements common.
2. Type

   d2wired.make
   

   This will make the executable in your $SCRATCH area.
3. d2wired.run executes the program. See the top of that script for details on the parameters. The script all.sh gives a few examples - those were the commands I used to create the WIRED demo. Control options are specified in the d2wired.tit title file. These can be overridden on the command line, eg.

   d2wired.run -t EK1318.65 IPRHAD=3 NEVMAX=10
   

   converts the first 10 hadronic events (by default it stops after one event) on VSN EK1318, file 65. Commented-out options in d2wired.tit tell you what the defaults are.

   The components to be written (TE, TK, and/or TV) are specified by setting INCLTE, INCLTK, and/or INCLTV=TRUE or FALSE. TEs show calorimeter energy deposits, TKs (not included by default) show the reconstructed charged tracks, and TVs show the charged tracks extrapolated to the vertex.

4. A list of created events is written to event.list (this is in PHDST format, where the first two numbers on each line give the run and event number). The output of each event is in a file rRRRRRR_eEEEEEE.asc , where RRRRRR is the run and EEEEEE is the event number (eg. r061832_e000074.asc). This naming scheme is required by WIRED.

Background

d2wired was written by Tim Adye, using ideas from Patrik Gunnarsson and Tord Malmgren's getevent routine (this can be compared with d2wired's equivalent, d2wput, by specifying IWRCOD=1) which they incoporated into DAFNE (see /afs/cern.ch/user/m/malmgren/cnvevt for their full setup). I chose to use SKELANA directly (which DAFNE also uses) in order to simplify the setup. The main improvements over cnvevt are, that d2wired

• reads all DST formats except for the miniDST. ShortDST, though, doesn't include the muon chamber hits (contrary to the shortDST documentation).
• shows secondary vertices. Only tracks unambiguously associated with a secondary vertex are offset. Other tracks are forced through the origin (this procedure needs improvement).
• offsets primary vertex position to the origin to simplify zooming (a better solution would be for WIRED to pick up an "initial translation" from the event file).
• shows jet assignments by colouring tracks and energy deposits. It might be necessary to adjust DJOIN (default is 5.0) for high energy events.
• excludes neutral "tracks" (neutral energy should be visible in the calorimeters).
• on the other hand cnvevt includes tracking detector hits (TEs). These were excluded from d2wired to simplify the display for non-experts.

The WIRED ASCII format was designed by Mark Dönszelmann and (I guess) Patrik Gunnarsson. The track and hit colours were defined later for ralwired97, so WIRED version 0.90 probably won't display them. Plans are afoot to move to an XML event description.

Possible Improvements

• Produce XML output once this format is defined.
• Improve vertex finding and fitting. Tracks should be fit to the vertex, rather than just translated. Unassociated tracks probably shouldn't be forced through the primary (perhaps don't extrapolate beyond their first-measured point).
• See whether we can get muon chamber information from the shortDST, perhaps by using track extrapolations (not perfect, but if the track is identified as a muon, then this can't be too far from the actual hit - hopefully good enough for a display).
• Create muon chamber hit-like objects for ralwired97, which doesn't display hits (TEs with zero energy). These could have a default energy, chosen to give a cube. Alternatively it may be possible to get ralwired97 to display hits (a configuration option?).
• Add options to include additional information (eg. tracking detector hits, particle id) in the output file. Most will require the XML event format. Mainly of use in the fully-featured version of WIRED.
• Add additional event selection options.
• Select default DJOIN (or other jet finding parameters) depending on the collision energy.
• Check that SAT, STIC, and/or VSAT information is displayed correctly. These are the luminosity monitors, but they can also be used as very-forward calorimeters.
• More ideas?