Software Minutes
 		       June 27, 2000


Pattern Recognition:
--------------------

Jim has sent the new version of the pattern recognition 
code to Art. Art started looking at the new code. The 
aim is to be able to deal with single track events 
(per time window) in the November run. Once this is done,
modifying the code for multiple tracks will be pursued.


First Guess:
------------

Art started looking at the first guess code written by
Roman to see how it should be modified to interface it
with the pattern recognition code. He expects that he
will need a couple of weeks for code modifying and 
debugging.


Kalman Filtering:
-----------------

The posting from the previous software meeting contained
a plot showing that the number of events successfully
reconstructed by the Kalman Filter (at a fixed "positron"
energy of 45 MeV) was angle dependent. While at 10 
degrees 96% of the events were successfully fitted, this
number went down to only 10% at 80 degrees. This was
not surprising as the covariance matrix was not optimized.

Since then I was able to confirm that the fraction of 
successfully reconstructed events depends on the choice 
of the covariance matrix. However, to be able to fit 100% 
of the events at all angles, I had to choose differnet 
starting values for the covariance matrix (depending 
on energy and angle). This is not a final conclusion,
however, and work is in progress to determine whether
something can be done to avoid changing the starting 
values of the covariance matrix. 

If this is not successful alternative approaches will 
be considered. One approach is to rely on the first 
guess to provide an approximate energy and angle and
choose the covariance matrix based on that. The second 
approach is to step through a range of values for the 
covariance matrix to find a good starting point. Both 
of these approaches are more cpu intensive (with the 
second likely more intensive than the first).


Event Display:
--------------

The event display has not been running in Texas. Paul
and Jim are now working on fixing this. They are also 
discussing the sort of modifications to the event display
to help view the results from the pattern recognition 
code.

We also discussed the types of modifications to the event
display that will be convenient to have for the November
run. Several technical questions came up which require a 
more focused discussion:

	- How do we display a "helical" track which 
	takes into account energy loss and multiple 
	scattering?
 
	- Do we want to display drift "circles" 
	or just the (u,v) results from fitting, and if 
	we want to display drift "circles" how do we
	handle the 3-D case and projections?


FORTRAN 90 Documentation:
-------------------------

Renee sent out email a couple of weeks ago pointing to
a web site that has a downloadable program which allows
documentation of FORTRAN 90 code. Bill Shin downloaded 
the script and started looking at it. The code is written 
in Perl scripting language. It understands FORTRAN 90 
structures etc. This script is designed to run through 
FORTRAN 90 code and produce a documentation of all the 
subroutines, structures, and pick all comment lines 
starting with "!!". The script also produces "html" files 
listing the subroutines. Each subroutine name may be 
clicked to display it's documentation.

A better understanding of what this script does will be 
achieved once the script is run on some FORTRAN 90 subroutines.
Bill will pursue this.


Data Base for Documenting Hardware Components:
----------------------------------------------

Renee and Andrei have been discussing the relational
data base for documenting the hardware components of
the E614 detector. Andrei has made good progress on 
this. His current development is on the web at

http://lin05.triumf.ca/~andr/e614db

Andrei will continue with this development. Once finished,
this page will be available to the people constructing 
the detector to document all hardware components.


AOB:
----

A new dual celera 533 will arrive at TRIUMF this week.
This machine will be used by Andrei for the summer, and
will then reside in the M13 counting room. The other
PCs that E614 owns so far are:

	- Dual pentium III 500
	- Dual pentium II 400
	- Dual pentium II 400

These machines will all be networked together. Glen
Marshal also volunteered his machine (dual pentium II 
400) to be connected to the network. Renee and Tracy 
started experimenting with networking linux machines.
Anyone who wants an account on this network should
contact Renee.

-- 
==================================================================
   Maher Quraan                        	Mail:                        
   office: Trailer A, TRIUMF	           TRIUMF
   email: quraan@triumf.ca                 4004 Wesbrook Mall
   phone: (604)222-1047 ext. 6492	   Vancouver, BC
   fax:   (604)222-1074                    Canada V6T 2A3

		http://www.thehungersite.com/

==================================================================

Created for the The Center for Subatomic Research E614 Project Projects Page.

Created by The CoCoBoard.