<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    Please find the minutes below and at
<a class="moz-txt-link-freetext" href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_April_27,_2016#Minutes">https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_April_27,_2016#Minutes</a><br>
    _______________<br>
    <br>
    <div id="globalWrapper">
      <div id="column-content">
        <div id="content" class="mw-body" role="main">
          <h1 id="firstHeading" class="firstHeading" lang="en"><span
              dir="auto">GlueX Offline Meeting, April 27, 2016 </span><span
              class="mw-headline" id="Minutes">Minutes</span></h1>
          <div id="bodyContent" class="mw-body-content">
            <div id="mw-content-text" dir="ltr" class="mw-content-ltr"
              lang="en">
              <p>There is a <a rel="nofollow" class="external text"
                  href="https://bluejeans.com/s/9xG1/">recording of this
                  meeting</a> on the BlueJeans site.
              </p>
              <p>Present:
              </p>
              <ul>
                <li> <b>CMU</b>: Naomi Jarvis, Curtis Meyer, Mike Staib</li>
                <li> <b>FSU</b>: Brad Cannon</li>
                <li> <b>JLab</b>: Mark Ito (chair), David Lawrence,
                  Paul Mattione, Dmitry Romanov, Nathan Sparks, Justin
                  Stevens, Simon Taylor, Beni Zihlmann</li>
                <li> <b>NU</b>: Sean Dobbs</li>
              </ul>
              <h3><span class="mw-headline" id="Announcements">Announcements</span></h3>
              <ol>
                <li> <b>GCC 4.9.2 on farm</b>. Mark went over <a
                    rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002307.html">his
                    email</a> describing the use of the <code>module</code>
                  command to set-up use of GCC 4.9.2 on the farm nodes.
                  The latest versions of our software have been built
                  with this compiler and installed at JLab.</li>
                <li> <b>Analysis Scripts Moved</b>. Paul led us through
                  <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002310.html">his
                    email</a> describing the move of plug-in template
                  generating scripts for use with the Analysis Library.
                  He received input from Nathan on the destination
                  location.</li>
                <li> <b>New git repository: hd_utilities</b>. Mark has
                  created a catch-all GitHub repository for
                  miscellaneous projects. See <a rel="nofollow"
                    class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002313.html">his
                    email</a> for details.</li>
                <li> <b>Bug fix release: sim-recon-1.11.1</b>. Mark
                  reminded us of the <a rel="nofollow" class="external
                    text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002314.html">recent
                    release</a> that fixes a now famous initialization
                  problem that affects tracking in multi-threaded jobs.</li>
                <li> <b>GlueX Software Help Email List</b>. Mark
                  reviewed <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002297.html">his
                    email</a> announcing the list. Collaborators are
                  encouraged to subscribe.</li>
              </ol>
              <h3><span class="mw-headline"
                  id="Review_of_minutes_from_April_13">Review of minutes
                  from April 13</span></h3>
              <p>We went over the <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_April_13,_2016#Minutes"
                  title="GlueX Offline Meeting, April 13, 2016">minutes</a>.
              </p>
              <ul>
                <li> Dmitry reported that the RCDB C++ API has been
                  built and tested on the ifarm (CentOS 6/5).</li>
                <li> Paul is re-doing his study of multi-threaded
                  behavior of hd_root with the fix to the famous
                  initialization problem.</li>
              </ul>
              <h3><span class="mw-headline" id="Calibration">Calibration</span></h3>
              <h4><span class="mw-headline"
                  id="Calibration_Train_Processing">Calibration Train
                  Processing</span></h4>
              <p>Sean brought us up-to-date. He showed a <a
                  rel="nofollow" class="external text"
href="https://halldweb.jlab.org/wiki/images/5/5e/Sdobbs_OfflineMtg_27Apr16.pdf">slide</a>.
              </p>
              <ul>
                <li> The last launch was done with the famous
                  initialization bug (FIB) present in the code. Sean
                  checked with the experts and the BCAL and FCAL
                  calibration did not suffer much from this.</li>
                <li> There were a few small problems with jobs.</li>
                <li> He will start soon with calibration of the data
                  from the last two weeks of the run. He will get the
                  timing calibrations done first.</li>
                <li> There were a few anomalies that arose around the
                  time of the magnet ramp to 1345 A. In particular the
                  tagger calibrations shifted by 0.5 to 1.0 ns. Also the
                  microscope shows a change in the difference between
                  FADC and TDC times where the change is a smooth
                  function of channel. Changes are roughly 3 to 4 ns.
                  This leads to bumps in the tagger vs. RF timing
                  distributions[?]. Alex Barnes is working on this
                  issue.</li>
                <li> We noted that the FIB was introduced on March 3.</li>
                <li> Sean is keeping his jobs at eight threads.</li>
                <li> He did some bench-marking on the code after the
                  JANA locking changes (those that Paul described at the
                  last meeting). He sees a 10-20% improvement in speed.
                  He did see instances in some calibration plug-ins
                  where there was extensive calculation being done while
                  holding on to a lock.</li>
              </ul>
              <h4><span class="mw-headline"
                  id="Time-of-Flight_Calibration">Time-of-Flight
                  Calibration</span></h4>
              <p>Beni has done a TOF calibration using the files from
                the calibration train. He sees 100 ps resolution per
                counter over the entire detector with some counters a
                bit better. The calibration is done by timing one plane
                against another, so for a given counter in one plane
                comparison can be done with multiple counters in the
                other plane. This allows for an easy visual consistency
                check of the quality of the calibration.
              </p>
              <h3><span class="mw-headline"
                  id="Spring_2016_Run.2C_Processing_Plans">Spring 2016
                  Run, Processing Plans</span></h3>
              <p>We had a wide-ranging discussion on plans for a
                comprehensive reconstruction pass on the recently
                acquired data.
              </p>
              <ul>
                <li> Sean reminded us that the plans were discussed at
                  the last Calibration Meeting. See item (3) of the <a
href="https://halldweb.jlab.org/wiki/index.php/April_20,_2016_Calibration#Minutes"
                    title="April 20, 2016 Calibration">minutes of that
                    meeting</a>. There is deadline of May 31 for all
                  calibrations to be complete so we can start the pass.</li>
                <li> An accurate estimate of the computing requirements
                  will come from statistics for the next Monitoring
                  Launch. We will not attempt to adjust account priority
                  on the farm until we have those estimates in hand.
                  That said, Sean thinks that the processing should take
                  about a month.</li>
                <li> We discussed tagging schemes and tag names for
                  identifying software and configuration parameters used
                  in the pass. Sean, Justin, Paul, and Mark all had
                  ideas. We did not come to conclusions and decided to
                  continue the discussion offline.</li>
                <li> Curtis wondered if we can get away with a smaller
                  number of files per run for Monitoring Launches. We
                  are now processing the first 20 files of all
                  "is_production" runs[?].
                  <ul>
                    <li> Sean thought that we might only need 2 to 5 or
                      so. He reported that this is enough to see ρs and
                      ωs. These signals depends on many parts of the
                      detector working together properly.</li>
                    <li> We noted that the faster the turn-around for
                      the launch, the faster problems can be found and
                      calibrations adjusted.</li>
                    <li> Paul will abort the current launch that has
                      just started. Sean will suggest to Paul a good
                      number of files to start up with again. The
                      current calibrations are good enough to learn
                      something. We do not have to delay further to
                      finalize calibrations for the last runs of the
                      run.</li>
                  </ul>
                </li>
                <li> We discussed the relative priority of producing
                  REST files vs. calibration/monitoring in the run up to
                  May 31. There was a consensus that we should
                  concentrate on the latter.</li>
              </ul>
              <h3><span class="mw-headline" id="GlueX_TTree_DSelector">GlueX
                  TTree DSelector</span></h3>
              <p>Paul led us through <a rel="nofollow" class="external
                  text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002311.html">his
                  email</a> describing a new ROOT selector, DSelector
                that specialized and improves the ROOT-supplied
                TSelector. It is described in detail on its own <a
                  href="https://halldweb.jlab.org/wiki/index.php/DSelector"
                  title="DSelector">wiki page</a>. It is meant to be
                used with ROOT trees produced by the Analysis Library
                and exploits knowledge of the reaction used to
                auto-magically provide reaction-specific operations,
                including PROOF support, saving work for the user. Brad
                has been using it recently and reports an excellent user
                experience. [Reminder: Paul pay Brad his $5]
              </p>
              <h3><span class="mw-headline"
                  id="Viewing_DEPICSvalue_with_hd_dump">Viewing
                  DEPICSvalue with hd_dump</span></h3>
              <p>David described a <a rel="nofollow" class="external
                  text"
href="https://mailman.jlab.org/pipermail/halld-offline/2016-April/002312.html">new
                  feature of hd_dump</a> that can greatly increase the
                speed of retrieval of EPICS information from the raw
                data when the program is run repeatedly. He relies on
                the native Linux file-caching mechanism to facilitate
                retrieval of data from memory rather than from disk.
              </p>
              <h3><span class="mw-headline" id="Transition_to_C.2B.2B11">Transition
                  to C++11</span></h3>
              <p>For the event-parsing work he is doing, David is
                incorporating a lot of C++11 language features. He is
                doing this on a branch, since these features are not
                supported by GCC 4.4. They will be supported when we
                move to 4.8 on June 1.
              </p>
              <p>He has noticed warning messages compiling our standard
                code (not his new stuff) with the new compiler that are
                not present with GCC 4.4. He was concerned that other
                people may be trying to fix these warnings as well and
                that would lead to duplication of effort and merge
                conflicts. There was not a lot of discussion, and we
                interpreted this as a lack of effort by developers to
                squelch warnings from compilers-of-the-future. We
                concluded that no group action is necessary at this
                time.
              </p>
              <p>Nathan asked about plans for going to ROOT 6. Mark
                thought that once we change to the new compiler, that
                step would be take-able. David thought we should wait a
                bit after June 1 so that we are not dealing with two
                issues at the same time. We agreed that that would be
                wise.
              </p>
              <h3><span class="mw-headline" id="RCDB_Status">RCDB Status</span></h3>
              <p>Dmitry updated on the status of RCDB. A lot a work has
                gone into it recently. See <a rel="nofollow"
                  class="external text"
href="https://docs.google.com/presentation/d/1_uhBIGhZOFy0v41ihW7RXVvonCy2LjrQ81dOAJ3FcFk/edit?usp=sharing">his
                  slides</a> for full details.
              </p>
              <ul>
                <li> The RCDB command line is now available. Simple
                  queries can now be done without having to write a
                  custom Python script. Dmitry showed some examples.</li>
                <li> An RCDB website is on the way.</li>
                <li> Simple plotting of values is now available.
                  Examples were shown. PDF files of the plots can be
                  created. Dmitry has a lot of ideas for future
                  sub-features.</li>
                <li> Custom definitions of "aliases" (set of logical
                  conditions to be satisfied) and "views" (lists of
                  variables to output) can now be defined by the user.
                  Definitions are kept in ~/rcdb_user_cfg. Dmitry showed
                  some examples. So now you can save your go-to queries
                  without having to type them in each time.</li>
                <li> He gave a list of items for future development.</li>
                <li> He is developing a facility for parsing the
                  configuration files so that individual parameters can
                  be extracted and turned into standard RCDB conditions.
                  <ul>
                    <li> If you have requests for parameters to extract,
                      contact Dmitry.</li>
                  </ul>
                </li>
                <li> Before he leaves (next week, yikes!) he plans to
                  release version 1.0. He listed the features supported
                  in this release.</li>
              </ul>
              <div class="printfooter">
                Retrieved from "<a dir="ltr"
href="https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_April_27,_2016&oldid=74810">https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_April_27,_2016&oldid=74810</a>"</div>
              <ul id="f-list">
                <li id="lastmod"> This page was last modified on 29
                  April 2016, at 17:55.</li>
              </ul>
            </div>
          </div>
        </div>
      </div>
    </div>
    <pre class="moz-signature" cols="72">-- 
<a class="moz-txt-link-abbreviated" href="mailto:marki@jlab.org">marki@jlab.org</a>, (757)269-5295
</pre>
  </body>
</html>