MUON ALGORITHMS GROUP REPORT			February and March 1999

The Mureco package has been split into three packages: muo_hit, muo_segment,
and muo_track. They have been officially released. These three packages, as
well as, muo_unpdata and muo_digi packages have been converted to ctbuild
packages.  Documents for muo_digi and muo_hit have been also updated.

Algorithms are being developed and tested for fast segment finding.  An 
algorithm using Neural Networks to find direction cosines of track segments in
the WAMUS PDT has been tested in an ideal case (three hits/layer and no 
noise).  The results are very promising.

Code for Pixel Geometry has been completed.(All other sub-systems except 
cosmic-caps are complete, debugged and tested.)  Some discrepancies for pixel 
geometry w.r.t. GEANT exist.  Pixels implemented in geometry have A layer
tilted, B and C layers flat.  There are cm size shifts between GEANT and
geometry, but problems are with GEANT.  We are working on fixes to the TZ
file and adding tilts to B and C layer pixels.

We are proceeding to implement B and C layer cosmic cap scintillators in 
geometry.  We have the data files.  They would be associated with the positions
of the chambers they are mounted on.  The A-phi classes will be modified to 
follow this convention.

Toroid material classes have been written, but positions need to be linked 
to those of the WAMUS chambers. 

Rick Jesik and George Alverson have been testing the MDT detector display.
George found some problems with geometry_system rotations. One bug has been
fixed, the other is due to a different convention from what one would expect - 
GeometryXform build method using Euler angles assumes active rotations but
Rotation class methods getPhi, getTheta, getPsi assume passive ones.  We have
a patch for this now.  After this, George was able to display MDT's correctly.