Release notes for the Multiplicity reconstruction package released March 16, 1999:

In this release, TPhMapSets have been integrated directly into the TPhMul class, replacing arrays of TPhPadMap pointers. These changes should not break any existing macros. New graphics getters allow you to plot directly from TPhMapSets in TPhMul or read from DST files. For example, if you have used runmul.C and writemul.C to create a DST called muldst.root, then in a later Phat session you could type

TFile ff("muldst.root")

.L $PHATHOME/phatmul/macros/plotmaps.C

plot((TPhMapSet *)ff.Get("e9disp"))

and follow prompts, to show the standard 4-frame displays. You can perform any other legitimate operation on a TPhPadMap extracted from a TPhMapSet, for example,

TPhMapSet *m1 = ff.Get("e9disp")

m1->GetMap("*-1")->GetIntegral(13)

gives you the reconstructed multiplicity for the 10th event in the DST. (In principle you can chain these statements, but CInt doesn't seem to like that, and in any case it's clearer to separate them.) Of course there are many ways to simplify them, although the casting (TPhMapSet *) is built into the ROOT TFile class and can't be avoided. These operations will eventually become scripts or ancillary class methods activated by pushbuttons; I've included them here only to help explain TPhMapSets.

The macro runmul.C now has a leak test mode, which you can activate by setting reportLevel to 5 in a copy of runmul.C; this will display the memory used by every object in your process [exclusive of other users] at the end of each event. To activate this you need also, as noted in runmul.C, to add the following line to a file called .rootrc in your home directory:

*.Root.ObjectStat: 1

Thanks to Patrick Decowski for suggesting this use of gObjectTable.

These changes will eventually find their way into the never-quite-current User's Guide to Multiplicity Reconstruction

The release has been tested under Linux (ppcln.mit.edu), OSF1 (abacus.mit.edu) and SunOS (phobos.nsrl.rochester.edu);

see http://web.mit.edu/Verdier/Public/MulStatus.htm for the most recent information. I will test the versioning reading MDC2-format DST's shortly.

Class changes:

TPhDispDefV.h: Correct defaults for unused displays

TPhMapSet.cxx and .h: Plot interfaces updated, dangerous fDDD and fDSD fields removed

TPhMul.cxx and .h: replaced arrays of TPhPadMap pointers with TPhMapSets; add graphics Get methods for mapsets

TPhPadMap.cxx and .h: Change plot interfaces to use TPhMapSets

TPhSensDef.cxx: Add getters for thickness and distance from origin, anticipating more precise survey results

Macro changes:

plotmaps.C: New plot facility works directly from TPhMapSets.

runmul.C: Add leak test facilities, use TPhVertex validity check

writemul.C: Write TPhMul's TPhMapSets, instead of creatng copies as in MDC2. Writing a subset of maps is planned for the next release.

stdmul.C: New example of a standard user output interface

Please send me corrections and problem reports.

You can return here to the Phobos Detector pages.

Last Edited: Tuesday, March 16, 1999 03:20:51 PM by Robin Verdier

Deus stanii regnet