TWIST SOFTWARE MEETING MINUTES
                          November 6, 2000


1. Efficiency code: 

The efficiency code has been written and tested. It will be available 
in time for the 3 module test. The code uses reconstructed tracks (fitted 
to wire centers) to find the cell/region were the hit is expected in 
each plane and then looking for a hit in that region. This gives a reasonable 
plane-by-plane efficiency. However, the accuracy of the cell-by-cell 
efficiency calculation is limited due to the coarse resolution. Depending 
on how the counting is done in this case, the cell-by-cell efficiency is 
either underestimated or overestimated. This will remain a problem until 
the timing is put in. For purposes of  setting up the HV, plane-by-plane 
efficiencies should be sufficient.

The Monte Carlo does not currently have efficiencies installed. The form
in which efficiencies should be installed in the MC was debated. It was
suggested that eventually we are likely to have to do an iterative processes,
were the thresholds are entered and the resulting efficiency is observed
until the process converges. It was agreed, however, that for the time being
a simple efficiency will be hard wired in the MC to allow for code testing.
Valpo volunteered to look into this.

2. Preparations for the November run: 

Sun-Chong will prepare the geometry file and the TDC map files needed. 
Maher will modify CFM and test MOFIA with the new geometry and map files.
Renee reported that a linear fit is sufficient to determine the threshold
setting at the VTX input (control point) vs the actual threshold (at the 
VTX ouput). 

3. Installing multiple scattering and energy loss in the Kalman code:

Sun-Chong has installed EL and MS in the code, and tested the code using
by generating positrons in the MC at energies of 10 - 40 MeV. An improvement
in the fit was observed, however, since there is no timing in the current
fitting code, it is hard to quantify this improvement.

4. Event Display:

Jim modified the event display to be able to display hits as grouped in the
pattern recognition code. The modifications work fine.

5. A new version of MOFIA:

A new version of MOFIA will be tagged as soon as the code settles down. A 
stable version of the pattern recognition code is needed. This will be 
discussed when Jim arrives at TRIUMF later this month. The new tagged version
will likely be ready sometime in December.

6. Geometry info:

Dave pointed out that some geometry information (that is not currently used
in the analysis) is wired into the MC and should be taken out and put into
the geometry file. Dave will pursue this further.

7. GEANT banks:

The plan is to have the following banks:

	- TDC banks.
	- An ascii bank for the FFCARDS.
	- An integer bank containing event information
	 (random# for event, event #, etc).
	- A real bank containing track information
	 (momentum, vertex of primaries and secondaries, etc.)

Maxim previously created the last two banks as one integer bank, but it was 
then suggested that he splits them into one real and one integer bank. This
resulted in a bug that Maxim is currently working to fix. The FFCARDS ascii
bank is still to be added.

8. Installing MC and MOFIA at U of A:

Problems have been encountered with installing the code at U of A. It was
agreed that we should decide on a LINUX OS version number, an ABSOFT compiler
version number and a version number of the CERN libraries that everyone should 
use. In the past, this has been a problem since we were using lin01 which is a 
TRIUMF machine, and had to adjust to new OS's asey are installed. Currently
we have a LINUX TWIST cluster at TRIUMF. The new ABSOFT version currently used
at TRIUMF should be distributed to all institutions. Art and Renee will follow
up on these points.

9. GEANT CPU:

Peter Gumplinger pointed out that the most CPU time savings are likely to come 
from modifying the geometry code. Peter will look into this. Also, flags should
be added to all the diagnostic sections so they are turned off in production runs.
Better yet, compiler flags should be installed to avoid repeated tests with IF
statements. Rob Macdonald will continue working on the digitization modifications
to improve the cpu time there.


-- 
==================================================================
   Maher Quraan                        	   TRIUMF               
   TRIUMF TWIST Collaboration	           http://www.triumf.ca
   email: quraan@triumf.ca                 4004 Wesbrook Mall
   phone: (604)222-1047 ext. 6333	   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.