<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Folks,</p>
    <p>Please find the minutes <a moz-do-not-send="true"
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_October_30,_2018#Minutes">here</a>
      and below.</p>
    <p>  -- Mark</p>
    <p>_________________________________________</p>
    <p>
    </p>
    <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, October 30, 2018, </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>Present:
              </p>
              <ul>
                <li> <b> CMU: </b> Hao Li, Reinhard Schumacher</li>
                <li> <b> FSU: </b> Sean Dobbs</li>
                <li> <b> JLab: </b> Ashley Ernst, Mark Ito (chair),
                  David Lawrence, Simon Taylor, Beni Zihlmann</li>
                <li> <b> W&M: </b> Justin Stevens</li>
              </ul>
              <p>There is a <a rel="nofollow" class="external text"
                  href="https://bluejeans.com/s/xuYcn/">recording of
                  this meeting</a> on the BlueJeans site. Use your JLab
                credentials to access it.
              </p>
              <h3><span class="mw-headline" id="Announcements">Announcements</span></h3>
              <ol>
                <li> <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2018-October/003403.html">New
                    release of halld_sim: version 3.6.0</a>. Released
                  October 22. Has changes from Colin Gleason for
                  amplitude-based generators.</li>
                <li> <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2018-October/003405.html">New
                    bug-fix release: halld_sim 3.1.1</a>. A special
                  release to get genEtaRegge going.</li>
              </ol>
              <h3><span class="mw-headline"
                  id="Review_of_minutes_from_the_October_16_meeting">Review
                  of minutes from the October 16 meeting</span></h3>
              <p>We reviewed <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Software_Meeting,_October_16,_2018#Minutes"
                  title="GlueX Offline Software Meeting, October 16,
                  2018">the minutes</a>.
              </p>
              <ul>
                <li> <b>Computing Review</b>. Curtis Meyer has
                  circulated a list of slide titles he proposes for the
                  Hall D presentation. David gave feedback. Now only
                  adding contest remains. </li>
              </ul>
              <h3><span class="mw-headline"
                  id="Monitoring_launch_ver_18_.40_NERSC">Monitoring
                  launch ver 18 @ NERSC</span></h3>
              <p>David reviewed <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld/2018-October/000645.html">his
                  recent email</a>.
              </p>
              <ul>
                <li> The monitoring launch at NERSC is mostly finished.
                  There are a few jobs where the 271 thread finished
                  successfully, but the 272nd won't. This seems to
                  happen with particular input files. These jobs were
                  tried on the gluons and ran without a problem.
                  However, recall that they ran on KNL architecture at
                  NERSC. David is requesting an interactive node at
                  NERSC to do a test.</li>
                <li> Bryan Hess has found a few more places in the
                  network where the MTU had room for increase. He is
                  curious about how the new configuration will perform.</li>
                <li> The problem with missing cache files causing jobs
                  to hang remains, likely due to deletion from the
                  auto-cache cleaner. David will pursue solutions with
                  Chris Larrieu.</li>
                <li> David is thinking about what to do next. Sean is in
                  favor of a monitoring launch using 10 files per run
                  dispersed throughout the Spring 2018 run. David told
                  Bryan that he is thinking of firing up another launch
                  at the end of next week.</li>
                <li> Beni asked about tape robot problems. David has not
                  seen any lately, but heavy tape use during data taking
                  may not have occurred recently.</li>
                <li> David and Mark will discuss the version XML file
                  that David is using on NERSC.</li>
              </ul>
              <h3><span class="mw-headline"
                  id="What_I_learned_at_BRNMW_2018">What I learned at
                  BRNMW 2018</span></h3>
              <p>David attended the Basic Research Needs for
                Microelectronics Workshop last week. Turns out he
                learned quite a lot, from topics like DNA memory
                technology to others like hotel meeting room door sound
                standards. Please see <a
                  href="https://halldweb.jlab.org/wiki/images/f/f8/20181030_BRNMW2018.pdf"
                  class="internal" title="20181030 BRNMW2018.pdf"> his
                  slides</a> for the details.
              </p>
              <h3><span class="mw-headline"
                  id="Events_lost_due_to_vertex_fit_for_Lambda">Events
                  lost due to vertex fit for Lambda</span></h3>
              <p>Hao gave us a detailed look at the problem and proposed
                solutions to the issue where kinematic fits with vertex
                constraints do not converge. See <a rel="nofollow"
                  class="external text"
href="https://halldweb.jlab.org/DocDB/0038/003812/001/Offline_software_meeting_Oct_30_HaoLi.pdf">his
                  slides</a> for the details.
              </p>
              <ul>
                <li> He described the channel where the problem seems
                  especially severe, γp→ΛΛ̅p.</li>
                <li> He sees a deficit in the π<sup>−</sup> lab polar
                  angle at around 40 degrees, where the yield drops to
                  nearly zero.</li>
                <li> The fits fail when the ROOT linear algebra package
                  encounters a singular matrix</li>
                <li> Two solutions were tried
                  <ol>
                    <li> Multiply all matrix elements by 10,000 before
                      testing the determinant, rescale results back down
                      as appropriate.</li>
                    <li> Use tolerance tuning: change the tolerance ROOT
                      uses to call a determinant "zero". This tolerance
                      is user settable.</li>
                  </ol>
                </li>
                <li> Both methods recover much of the deficit, with
                  tolerance tuning showing better results.</li>
                <li> The fix has been put in a branch (kinFitter_debug)
                  of halld_recon so other can try it.</li>
              </ul>
              <p>From the discussion:
              </p>
              <ul>
                <li> Mark worried that the input covariance matrix for
                  the measured kinematic quantities may have errors.
                  that lead to this problem. Fixes that give fit
                  convergence may mask a deeper problem.</li>
                <li> Reinhard has done some research on the subject.
                  Other experiments have encountered similar problems.
                  Even if there are problems with the covariance matrix
                  formation, the proposed fixes should remain in place.
                  They give us a more robust approach to kinematic
                  fitting in general.</li>
                <li> Reinhard also pointed out that having others try
                  out the fix would answer concerns about the effect on
                  execution time and precision of results.</li>
                <li> Justin suggested trying the branch in the context
                  of an analysis launch as a test.</li>
              </ul>
              <h3><span class="mw-headline"
                  id="CCDB_versions_and_SQLite_file_management">CCDB
                  versions and SQLite file management</span></h3>
              <p>Mark led a discussion prompted by a incident where the
                CCDB had errors in one of its calibration sets (a
                translation table). From the agenda:
              </p>
              <ul>
                <li> Communication: folks should write to the <a
                    rel="nofollow" class="external text"
                    href="https://groups.google.com/forum/#%21forum/gluex-software">Software
                    Help List</a> when things are broken so others can
                  avoid problems.</li>
                <li> Finding CCDB SQLite backups: to make it easier to
                  find the backup versions we could create links like
                  "ccdb_current", "ccdb_yesterday", and/or
                  "ccdb_last-week".</li>
                <li> Move backups to a more obvious place: We could move
                  the backup version into the same directory as the
                  latest version, i. e. <a rel="nofollow"
                    class="external text"
                    href="https://halldweb.jlab.org/dist/">/group/halld/halldweb/www/html/dist</a>,
                  which is web-accessible.
                  <ul>
                    <li> We could reorganize that directory. There is a
                      lot of heterogeneous stuff there. Any re-org would
                      break old webpage links and scripts however.</li>
                  </ul>
                </li>
              </ul>
              <p>There was actually not a lot of discussion to lead.
                People with strong feelings should contact Mark.
              </p>
              <h3><span class="mw-headline"
                  id="Review_of_Offline_Work_Packages">Review of Offline
                  Work Packages</span></h3>
              <p>Mark update the list of packages based on discussion
                from the last meeting. We took a look at <a
                  rel="nofollow" class="external text"
href="https://halldweb.jlab.org/wiki-private/index.php/2018_Work_Packages">revised
                  list</a>. Items now appear in two lists: "Analysis
                Software" and "Software Infrastructure". Only the
                assignment of volunteers remains to be done.
              </p>
              <h3><span class="mw-headline"
                  id="Review_of_recent_issues_and_pull_requests">Review
                  of recent issues and pull requests</span></h3>
              <h4><span class="mw-headline" id="halld_recon">halld_recon</span></h4>
              <ul>
                <li> <a rel="nofollow" class="external text"
                    href="https://github.com/JeffersonLab/halld_recon/issues/40">Issue
                    #40: Sean broke the tracking again...</a> Problem
                  seen with ρ yield. Sean is looking at it.</li>
                <li> <a rel="nofollow" class="external text"
                    href="https://github.com/JeffersonLab/halld_recon/issues/30">Issue
                    #30: Trigger monitoring</a> Sean has ideas to help
                  identify LED triggers that get flagged as physics.</li>
                <li> <a rel="nofollow" class="external text"
                    href="https://github.com/JeffersonLab/halld_recon/pull/43">Pull
                    Request #43: fix NAN in return value of walk
                    correction code. Make sure that square root take a
                    positive value larger than zero</a> Beni eliminated
                  problem when the ADC amplitude is non-positive.</li>
              </ul>
              <h4><span class="mw-headline" id="halld_sim">halld_sim</span></h4>
              <ul>
                <li> <a rel="nofollow" class="external text"
                    href="https://github.com/JeffersonLab/halld_sim/issues/16">Issue
                    #16: mcsmear crash</a> There is a problem when RCDB
                  cannot find a file for two runs from Fall 2016. Sean
                  filed this as an RCDB issue.</li>
              </ul>
              <h3><span class="mw-headline"
                  id="Review_of_recent_discussion_on_the_GlueX_Software_Help_List">Review
                  of recent discussion on the GlueX Software Help List</span></h3>
              <p>We went over <a rel="nofollow" class="external text"
                  href="https://groups.google.com/forum/#%21forum/gluex-software">the
                  list</a>. No discussion of note.
              </p>
            </div>
            <div class="printfooter">
              Retrieved from "<a dir="ltr"
href="https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_October_30,_2018&oldid=89946">https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_October_30,_2018&oldid=89946</a>"</div>
          </div>
        </div>
      </div>
      <div id="footer" role="contentinfo">
        <ul id="f-list">
          <li id="lastmod"> This page was last modified on 2 November
            2018, at 17:14.</li>
        </ul>
      </div>
    </div>
    <pre class="moz-signature" cols="72">-- 
Mark Ito, <a class="moz-txt-link-abbreviated" href="mailto:marki@jlab.org">marki@jlab.org</a>, (757)269-5295
</pre>
  </body>
</html>