From: Michel Jaffre [jaffre@lal.in2p3.fr]
Sent: Friday, January 26, 2001 4:11 AM
To: Harry Melanson
Cc: Amber Boehnlein; Heidi Schellman; Ron Madaras; Frederic Fleuret;
Florian Beaudette; Melissa Ridel
Subject: Re: Status reports at Algorithms meeting

Harry,

 this is the changes which have been implemented ih the EMId packages


 semreco :

  Florian Beaudette has implemented the "road method", the
SEMparticleChunk will contain the list of channels in the road
for 3 values of the dismerge parameter.
 The list contains the channels in the larger road and channels are
tagged according to their inclusion in road 1,2 or 3

 No selection is however based on the estimators computed from these
roads.
 So the only "calorimeter selection" is still from the association
track-towers requiring 80% of em energy in the associated towers.

 Frederic did also some clean-up concerning the vertex selection
and also replaced cout messages by ErrorLogger messages

cellNN :

 Melissa has corrected a bug during the construction of the 3D cluster
from the floor clusters which resulted in the c.o.g to be outside the
calorimeter in heavy occupancy events. The consequence was the
"famous" coreE=0.0 in EMReco-cnn algorithm.

emreco :

 Although the coreE=O. is thought to be cured, the code is now protected
and an ErrLog will appear.

 Also some cout warnings have been replaced by ErrorLogger 
messages.


Michel 


*******************************************************************
*       Universite Paris-Sud       phone: +33 (0)164468926        *
*       LAL     B.P. 34              fax: +33 (0)164468546        *
*       F-91898 ORSAY Cedex        email:  jaffre@lal.in2p3.fr    *