Re: Vertex

From: Christian Holm Christensen (cholm@hehi03.nbi.dk)
Date: Thu Nov 09 2000 - 06:51:09 EST

  • Next message: Anders Holm: "Re: Vertex"

    Hi Bjorn et al, 
    
    On Wed, 8 Nov 2000 13:34:35 -0500 (EST)
    Bjorn H Samset <bjornhs@rcf.rhic.bnl.gov> wrote
    concerning ": Re: Vertex":
    > Yes, but this is not necesearily meant to last. The module was actually
    > "released" a bit prematurely since this part is not really ready, but I
    > did it to make resolution-investigations etc. a bit easier.
    
    I realise that the vertex algorithms is are work in progress, and
    ofcourse the "goodness" criteria will evolve as further work is put
    into it. My previous mail in this thread was merely an attempt to get
    some discussion going on how to determine the "best" vertex in a given
    event. 
    
    > 1) The ZDC's have a resolution sigma of 7-9 cm. This is so wide that it
    > would kill any fine res. needed esp. for PID in FS.
    
    I had a short discussion with Ian yesterday, over which I realiased
    that I probably wasn't clear enough on what I meant with "weighted
    mean". 
    
    What I imagine one could do, was to assign some "confidence factor" to
    each of the four algorithms, ofcourse based on scrutinising
    investigations, and the "best" vertex could be some linear combination
    of the found verticies. 
    
    In the extreem limit, this confidence factor is just a cut. 
    
    Of course, as Bjorn points out, the correlation of the vertex found by
    the cluster and the track algorithm, needs to be addressed. 
    
    > I have been looking through the output from BrTPMClusterVertexModule
    > today, and I believe I can easily remove at least part of this by
    > using one or two more internal cuts in the module 
    
    Please PLEASE document these cuts, and allow the user to change them. 
    
    As a general comment, I'd like to direct peoples attention to the nice
    features for documenting your class directly in the code, as provided
    by the THtml class in ROOT. What I do, is to prepare some document in
    LaTeX (so that I may have fancy maths stuff in it), and then translate
    that into HTML using LaTeX2HTML. This document I past into the source
    code:
    
       //___________________________________________________________________
       //Begin_Html
       /*
       </pre> <!-- needed since we're inside a <pre> block -->
           <!-- Here goes the HTML document from LaTeX2HTML -->
       <pre> <!-- back to the <pre> block -->
        */
       //End_Html
    
    You'll probably have to include some GIFs in your document, for maths
    and the like. These could be added to the CVS repository in a common
    "gif" directory. 
    
    Again, I'd like to urge someone at BNL to regularly update the HTML
    pages. I can do it froim here, but since you need an X connection to
    do it, it would be more sensible for someone with a high-speed X
    connection (i.e., not across the atlantic via SSH) to do that. In
    ~brahmlib/bin/ (I can't remember the name) I put a script that'll
    semi-automate the process. 
    
    A few liras(!) worth. 
    
    Yours, 
    
    Christian  -----------------------------------------------------------
    Holm Christensen                             Phone:  (+45) 35 35 96 91 
      Sankt Hansgade 23, 1. th.                  Office: (+45) 353  25 305 
      DK-2200 Copenhagen N                       Web:    www.nbi.dk/~cholm    
      Denmark                                    Email:       cholm@nbi.dk
    



    This archive was generated by hypermail 2b29 : Thu Nov 09 2000 - 06:51:47 EST