The structure of the LCT data n-tuple is described below.  Before you
start I recommend that you get the postscript files from
 
        http://ucdcms.ucdavis.edu/www/cms_eta.html
        http://ucdcms.ucdavis.edu/~fisyak/cms/blpict/me.html
 
and print them for your reference along with this email message.  Then
run paw and load your data file containing ntuple 500.
 
    First note that each entry in the ntuple represents the LCTs produced in 
an entire endcap muon measurement station per event. So, for example, if you 
shoot a single muon through the endcap you should get four LCTs most of the 
time.  The first four entries in your ntuple, then,  will be from the same 
event.  You can keep track of all the LCTs from a single event using the  first
n-tuple data field, IDEV.

In paw, try the following command.
 
       PAW> nt/scan 500 idev=1 ! ! ! idev idtype nstlct nwrlct
 
For my simple test events that unrealistically contain a single muon track,
I see the following.
 
 
For my simple test events that unrealistically contain a single muon track,
I see the following.
 
+-------+-------------+-------------+-------------+-------------+
| Event |   IDEV      |   IDTYPE    |   NSTLCT    |   NWRLCT    |
+-------+-------------+-------------+-------------+-------------+
|     1 |  1          |  31212      |  1          |  1          |
|     2 |  1          |  31222      |  1          |  1          |
|     3 |  1          |  31232      |  1          |  1          |
|     4 |  1          |  31242      |  1          |  1          |
+-------+-------------+-------------+-------------+-------------+
 
Notice that I've only included the single dimension n-tuple data fields in
the command.  In this table, "Event" is a misnomer.  It really should be
"ntuple entry".  Try this for other events (IDEV=2, 3, 4, ...) and you 
 
     IDTYPE refers to the endcap muon measurement station id used
by GEANT and is always of the form 312nm for the muon endcap, where n is the
station position in z, and m is the ring number in r.  If you really want to
see the meaning of IDTYPE explicitly, find a picture of the CMS detector
( at http://ucdcms.ucdavis.edu/www/cms_eta.html if all else fails ) and then
try the following.
 
        PAW> null 0 1200 0 1200
        PAW> nt/pl 500.sqrt(xglmat**2+yglmat**2)%zglmat idtype=31210 ! ! ! s
        PAW> nt/pl 500.sqrt(xglmat**2+yglmat**2)%zglmat idtype=31212 ! ! ! s
        PAW> nt/pl 500.sqrt(xglmat**2+yglmat**2)%zglmat idtype=31213 ! ! ! s
        .
        .
        .
          < and so on... for each of the IDTYPE values >
 
NSTLCT and NWRLCT are the total number of strip LCTs and total number of 
wire LCTs respectively for that particular muon station and event.  So in the 
table above, my single muon has produced one strip and one wire LCT in the 
four
stations that it passed through.  If instead I had shot two muons at the
same eta, but with possibly different phi's then there would still be only 
four entries but NSTLCT and NWRLCT would be 2 instead.  Does this make sense?
 
     Problem is, there are numerous individual trapezoidal chambers within
a muon station refered to by IDTYPE.  First, for each IDTYPE, there is a 
chamber located at both +z and -z.  Second, for each IDTYPE and z position,
there are up to 36 chambers arranged in phi with an additional z offset to
allow overlaps ( see http://ucdcms.ucdavis.edu/~fisyak/cms/blpict/me.html ).
If you want to decide, therefore, whether an individual trapezoidal chamber 
has been hit twice when NSTLCT=2, you need more information than just IDTYPE.
To find the exact chamber you need to decode the NSNVS1, NSNVS2 and NSNVS3
ntuple data fields.
 
        NSNVS1 tells which end [ 1=forward (+z), 2=backward (-z) ]
        NSNVS2 tells which overlap [1=closest to i.p., 2=farthest from i.p.]
        NSNVS3 tells which phi-ordering [ ranges from 1-18 ]
 
Note that you get the 36 total chambers in a forward ring by combining
NSNVS2 with NSNVS3.  Once again, to see this explicitly, try
 
        PAW> nt/pl 500.xglmat%yglmat%zglmat idtype=31221.and.nsnvs2=1
        PAW> nt/pl 500.xglmat%yglmat%zglmat idtype=31221.and.nsnvs2=1 ! ! ! s
 
with all the combinations of idtype and nsnvs until you get the idea.  Once
you've understood the thing presented so far you should have a pretty good
idea how the chamber volumes are represented in the ntuple.
 
 
     The next thing is the details of the LCTs themselves.  The following
table lists the definitions of each entry in ntuple 500.  There is a section
for both strip and wire LCT data.  I hope this helps.  Ask me further
specific questions if you need.
 
       Cheers,
                Jeff Rowe.
 
-------------------------------------------------------------------------------
 
**********
 CSC LCTs
**********
 Name    *
**********
IDEV      Event # (e.g. all entries with IDEV=1 come from event #1)
IDTYPE    Muon measurment station id
NSTLCT    Number of strip LCTs per event per measurement station
NSTLCT    Number of wire LCTs per event per measurement station
 
       The following refer to the details of the i-th strip LCT
 
LSTPAT(i) is the LCT pattern ( see mc_str_road.F for the definition )
PHILCT(i) is the local phi position
PHIGLB(i) is the global phi position
PASANG(i) is the local bending angle defined by the LCT hit pattern
IHSLCT(i) is the half-strip id hit in the key layer
TSTLCT(i) is the time of the LCT hit
NHTLST(i) is the number of hits in LCT pattern (should always be 4,5 or 6)
LCTSGN(i) is the signed charge of the track (0=+, 1=-)
NSNVS1(i) tells whether the chamber is in the forward or backward region
NSNVS2(i) tells the chamber offset
NSNVS3(i) tells the phi-ordering of the chamber
 
       For the following, the HITS bank is searched for the simulated hit
closest to the LCT.  The resulting matched hit positions are saved here and
represent the "true" position of the track that passed through the chamber.
Note that none of these matched values will be known in the real experiment.
 
PHIMAT(i) is the phi-position of the matched hit
XGLMAT(i) is the x-position of the matched hit
YGLMAT(i) is the y-position of the matched hit
ZGLMAT(i) is the x-position of the matched hit
TIMMAT(i) is the time of the matched hit
PXXMAT(i) is x-componant of the unit vector in the direction of the momentum
PYYMAT(i) is y-componant "   "   "
PZZMAT(i) is y-componant "   "   "
ABPMAT(i) is absolute value of the momentum of the matched track
 
       The following refer to the details of the j-th wire LCT. Note that
there is no correspondance between the strip LCTs and wire LCTs (i.e. strip 
LCT #1 and wire LCT #1 don't neccisarily come from the same track).  
 
LWRPAT(j) is the LCT pattern ( see mc_wir_road.F for the definition )
ETALCT(j) is the eta position of the LCT
IWGLCT(j) is the wire-group hit in the key layer
TWRLCT(j) is the time of the 2nd hit in the LCT
NHTLWR(j) is the number of hits in the LCT pattern
NWNVS1(j) |
NWNVS2(j)  > Are the same as above but for wires instead.
NWNVS3(j) |