--------------514B23D3321C
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Hello,
I have added documentation to the file ecfph.F, one of the major parts
of the EC reconstruction code. EC experts, please read and comment.
- Will
--------------514B23D3321C
Content-Type: text/plain; charset=us-ascii; name="temp.dat"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="temp.dat"
SUBROUTINE ECFPH(L,S)
c
c _begin_doc
c RCS ID string
c $Id: ecfph.F,v 1.7 1997/12/23 00:06:44 stepanya Exp $
c
c Documentation for subroutine ECFPH
c
c Purpose: The name of the subroutine means 'EC - fit peaks to hits'.
c This routine is called only by ecfit_analyze, after ecfsp has been called
c for each view (u,v,w) of the current layer. Therefore the initial guess
c for the peaks in each view has been defined, and the present routine tries
c to match them together into 'hits'.
c
c First there is a threefold loop which looks at all possible combinations
c of peaks in the u, v, and w views. For each iteration of the loop, the
c 'dalitz' condition is checked using the centroids in each of the three
c views, using the RMS as an estimate of the uncertainty in each view. A
c limit is imposed on the number of possible hits by the parameter MAXpsble,
c currently defined as being equal to MAXhits, which is currently defined in
c Ec_general.PAR as being = 10. If this limit is exceeded, the event is not
c analyzed.
c
c If the 'dalitz' condition is satisfied (geometrically plausible event),
c the path from the hit position along the strip is calculated, and the
c path length, identity of the peaks, and identity of the hit is stored and
c the threefold loop ends.
c
c Now that the path lengths are known for all three views, the routine
c ecfsp is called a second time, with iterr=2. This re-calculates the peak
c characteristics after doing an attenuation length correction, so that
c in principle a more accurate centroid, rms and higher moments are
c obtained.
c
c Next, peaks which are part of more than one hit are treated. If for all
c hits there is only one peak in each view, then the routine proceeds to
c the next step. If a given peak contributes to multiple hits, then the
c energy in each hit due to that peak is calculated as being proportional
c to the relative sizes of the multiple hits as measured in other views.
c That is, if there are two hits, both of which have the same 'u' peak,
c the energy in v and w is added for each of the hits, and the ratio of
c these summed energies determines how much of the u peak's energy is
c assigned to each of the two hits. The code does not attempt to handle
c more complicated events, which, of course, are topologically possible.
c
c Then the hit is again checked for the 'dalitz' condition, using the
c refined (attenuation-length-corrected) centroid and RMS values; at the
c present time no action is taken based on the results of this check.
c Next the hits are sorted by energy, and a threshold cut is applied. This
c threshold value is set by the tcl parameter 'EcFitEdge_hit_threshold',
c and it is a true energy threshold since attenuation corrections have
c already been applied. After this, a number of characteristics of the
c hit are stored, and the subroutine returns.
--------------514B23D3321C--