<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html;
      charset=windows-1252">
  </head>
  <body>
    <p>I'm not sure about the physics of the following WIEN2k 19.2
      parallel calculation (with all patches at [1] applied), but
      mechanically the "x qtl -p -telnes" seems to have run without
      error.</p>
    <p><br>
    </p>
    <p>I typically have SCRATCH in my .bashrc set to "./" but used
      another location "/home/username/wiendata/scratch" as seen below. 
      Does a simple k-point parallel calculation like the one below work
      on your system?  I haven't tried mpi parallel yet.  On the other
      hand, I have noticed a possible issue that if one forgets to setup
      a .machines file and tries to run a parallel calculation that
      qtlpara_lapw seems to fail switching over to the serial
      calculation mode as shown under [2] below.  If one compares for
      example lapw2para_lapw and qtlpara_lapw, as illustrated by [3]
      below, the qtlpara_lapw may be missing some additional code that
      could be needed to get that to work.<br>
    </p>
    <p><br>
    </p>
    <p>username@computername:~/wiendata/diamond$ grep SCRATCH ~/.bashrc<br>
      export SCRATCH=/home/username/wiendata/scratch<br>
      username@computername:~/wiendata/diamond$ ls<br>
      diamond.struct<br>
      username@computername:~/wiendata/diamond$ init_lapw -b<br>
      ...<br>
        init_lapw finished ok<br>
      username@computername:~/wiendata/diamond$ cat .machines<br>
      1:localhost<br>
      1:localhost<br>
      granularity:1<br>
      extrafine:1<br>
      username@computername:~/wiendata/diamond$ run_lapw -p<br>
      ...<br>
      in cycle 11    ETEST: .0001457550000000   CTEST: .0033029<br>
      hup: Command not found.<br>
      STOP  LAPW0 END<br>
      STOP  LAPW1 END<br>
      STOP  LAPW1 END<br>
      STOP LAPW2 - FERMI; weights written<br>
      STOP  LAPW2 END<br>
      STOP  LAPW2 END<br>
      STOP  SUMPARA END<br>
      STOP  CORE  END<br>
      STOP  MIXER END<br>
      ec cc and fc_conv 1 1 1<br>
      <br>
      >   stop<br>
      username@computername:~/wiendata/diamond$ cp
      $WIENROOT/SRC_templates/case.innes diamond.innes<br>
      username@computername:~/wiendata/diamond$ x qtl -p -telnes<br>
      running QTL in parallel mode<br>
      calculating QTL's from parallel vectors<br>
      STOP  QTL END<br>
      6.4u 0.1s 0:06.59 100.0% 0+0k 0+8024io 0pf+0w<br>
      username@computername:~/wiendata/diamond$ cat diamond.inq<br>
      0 2.20000000000000000000<br>
      1<br>
      1 99 1 0<br>
      4 0 1 2 3<br>
      username@computername:~/wiendata/diamond$ x telnes3<br>
      STOP TELNES3 DONE<br>
      3.3u 0.0s 0:03.39 99.7% 0+0k 0+96io 0pf+0w<br>
    </p>
    <p><br>
    </p>
    <p>[1] <a class="moz-txt-link-freetext" href="https://github.com/gsabo/WIEN2k-Patches/tree/master/19.2">https://github.com/gsabo/WIEN2k-Patches/tree/master/19.2</a><br>
    </p>
    <p><br>
    </p>
    <p>[2] Error when qtlpara_lapw tries to switch to single mode during
      "x qtl -p -telnes":<br>
    </p>
    <p><br>
    </p>
    <p>username@computername:~/wiendata/diamond$ cat .machine<br>
      cat: .machine: No such file or directory<br>
      username@computername:~/wiendata/diamond$ run_lapw -p<br>
      ...<br>
      in cycle 11    ETEST: .0001457550000000   CTEST: .0033029<br>
      hup: Command not found.<br>
      STOP  LAPW0 END<br>
      STOP  LAPW1 END<br>
      STOP  LAPW2 END<br>
      STOP  CORE  END<br>
      STOP  MIXER END<br>
      ec cc and fc_conv 1 1 1<br>
      <br>
      >   stop<br>
      username@computername:~/wiendata/diamond$ cp
      $WIENROOT/SRC_templates/case.innes diamond.innes<br>
      username@computername:~/wiendata/diamond$ x qtl -p -telnes<br>
      single: label not found.<br>
      0.0u 0.0s 0:00.01 0.0% 0+0k 0+0io 0pf+0w<br>
      error: command   /home/username/WIEN2k/qtlpara qtl.def   failed<br>
    </p>
    <p><br>
    </p>
    <p>[3] Grep difference between qtlpara_lapw and lapw2para_lapw:<br>
    </p>
    <p><br>
    </p>
    <p>username@computername:~/wiendata/diamond$ grep "single"
      $WIENROOT/qtlpara_lapw<br>
      testinput .processes single<br>
      username@computername:~/wiendata/diamond$ grep "single"
      $WIENROOT/lapw2para_lapw<br>
      testinput .processes single<br>
      single:<br>
      echo "running in single mode"<br>
    </p>
    <p><br>
    </p>
    <div class="moz-cite-prefix">On 10/20/2020 12:24 PM, Christian
      Søndergaard Pedersen wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:756c28d3e762403699bc2d90d7a9de56@dtu.dk">
      <meta http-equiv="Content-Type" content="text/html;
        charset=windows-1252">
      <style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;}</style>
      <div id="divtagdefaultwrapper" dir="ltr" style="font-size: 12pt;
        color: rgb(0, 0, 0); font-family: Calibri, Helvetica,
        sans-serif, "EmojiFont", "Apple Color
        Emoji", "Segoe UI Emoji", NotoColorEmoji,
        "Segoe UI Symbol", "Android Emoji",
        EmojiSymbols;">
        <p>Greetings</p>
        <p><br>
        </p>
        <p>I am trying to run qtl in order to calculate the partial
          charge densities for the telnes3 program. The following fails,
          generating the error in the subject line:</p>
        <p><br>
          <span style="font-family: "Courier New", monospace;"></span></p>
        <span style="font-family: "Courier New", monospace;"></span>
        <p><span style="font-family: "Courier New",
            monospace;">    run_lapw -p</span><br>
          <span style="font-family: "Courier New", monospace;"></span></p>
        <span style="font-family: "Courier New", monospace;"></span>
        <p><span style="font-family: "Courier New",
            monospace;">    x qtl -p -telnes</span><br>
        </p>
        <p><br>
        </p>
        <p>Meanwhile, the following works:</p>
        <p><br>
        </p>
        <p><span style="font-family: "Courier New",
            monospace;">    run_lapw -p</span></p>
        <span style="font-family: "Courier New", monospace;"></span>
        <p><span style="font-family: "Courier New",
            monospace;">    x lapw2 -p -qtl</span></p>
        <p><br>
        </p>
        <p>However, lapw2 does not generate the terms necessary for
          telnes3, which subsequently fails with error message "isplit
          needs to be 99".</p>
        <p>From reading this thread:</p>
        <p><a
href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg05792.html"
            class="OWAAutoLink" id="LPlnk477036" previewremoved="true"
            moz-do-not-send="true">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg05792.html</a></p>
        <p><span class="OWAAutoLink"><br>
          </span></p>
        <p><span class="OWAAutoLink">... I understand that it may be
            related to $SCRATCH. When I look inside lapw1.def and
            lapw2.def, I find:</span></p>
        <p><span class="OWAAutoLink"><br>
          </span></p>
        <p><span class="OWAAutoLink"><span><span style="font-family:
                "Courier New", monospace;">   
              </span><span style="font-family: "Courier New",
                monospace;">10,'/scratch/chrsop/case.vector',
                'unknown','unformatted',9000</span></span><br>
          </span></p>
        <p><span class="OWAAutoLink"><br>
          </span></p>
        <p><span class="OWAAutoLink">... while qtl.def contains the
            following:</span></p>
        <p><span class="OWAAutoLink"><br>
          </span></p>
        <p><span class="OWAAutoLink"></span></p>
        <div><span style="font-family: "Courier New",
            monospace;">     9,'/scratch/chrsop/case.vector',
            'unknown','unformatted',9000</span><br>
          <span style="font-family: "Courier New", monospace;">   
          </span><span style="font-family: "Courier New",
            monospace;">10,'/scratch/chrsop/case.vectordn',
            'unknown','unformatted',9000</span></div>
        <p><br>
        </p>
        <p>Note that lapw1 and lapw2 run smoothly. Presumably, qtl goes
          looking for case.vectordn, which is not generated during
          run_lapw. I tried deleting the offending line in qtl.def and
          running:</p>
        <p><br>
        </p>
        <p><span style="font-family: "Courier New",
            monospace;">    x lapw1 -p</span></p>
        <p><span style="font-family: "Courier New",
            monospace;">    x qtl -p -telnes</span><br>
        </p>
        <p><br>
        </p>
        <p>... but this failed for the same reason as before, and when
          the job finished, qtl.def once again had line 10 as shown
          above. My case.inq file looks like:</p>
        <p><br>
        </p>
        <div><span style="font-family: "Courier New",
            monospace;">    </span><span style="font-family:
            "Courier New", monospace;"><span
              style="font-family: "Courier New", monospace;">0
            </span>8.73846153846153846153</span><br>
          <span style="font-family: "Courier New", monospace;">   
            1</span><br>
          <span style="font-family: "Courier New", monospace;">   
            1 99 1 0</span><br>
          <span style="font-family: "Courier New", monospace;">   
            4 0 1 2 3</span></div>
        <br>
        <p>Any help in solving this matter would be greatly appreciated.<br>
        </p>
        <p><br>
        </p>
        <p>Best regards</p>
        <p>Christian<br>
        </p>
      </div>
    </blockquote>
    <blockquote type="cite"
      cite="mid:756c28d3e762403699bc2d90d7a9de56@dtu.dk">
    </blockquote>
  </body>
</html>