<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p><font face="Times New Roman">By the way, I suspect that the full
        convergence tests package might only do the k-point convergence
        test for a non-hf scf as adjusted by "x kgen".</font></p>
    <p><font face="Times New Roman">If your doing an hf calculation,
        does the test_conv-5.0.tar.gz have an hf feature that uses the
        run_kgenhf_lapw for the k-point convergence test for the scf [
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19131.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19131.html</a>
        ].<br>
      </font></p>
    <p><font face="Times New Roman">Also for the band structure k-point
        convergence test, is test_conv-5.0.tar.gz changing how fine the
        mesh is in case.klist_band (generated manually by hand,
        automatically by own script or XCrysDen?) before proceeding with
        run_bandplothf_lapw and "x spagehetti -hf" for an hf calculation
        [
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg18937.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg18937.html</a>
        ,
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19403.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19403.html</a>
        ].<br>
      </font></p>
    <div class="moz-cite-prefix">On 12/23/2019 1:33 AM, Gavin Abo wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:8f1c9cfb-872c-15a8-3961-ac70a441dce7@crimson.ua.edu">I
      don't recall seeing "test_conv-5.0.tar.gz" before.  I'm guessing
      that is the newer version of the full convergence tests package at
      [1] (unsupported at least by this group), which seems to not be
      accessible to the public (and likely only accessible to and
      perhaps supported by the private group [2]).
      <br>
      <br>
      I'm guessing the algorithm that gets the gap value from the
      valence band maximum (VBM) and conduction band minimum (CBM) is
      not accurate for your calculation case.  Therefore, I would most
      likely go with the gap that you get yourself from the raw data
      (case.spaghetti_ene [3] or [4]) of the band structure plot from
      spaghetti.
      <br>
      <br>
      If the differences are small like in [5], a percent difference [6]
      of about 2% [= (1.54 eV - 1.51 eV)/{(1.54 eV + 1.51 eV)/2}*100%]
      seems acceptable.  I believe greater than about 5% in when you
      should be concerned [7].
      <br>
      <br>
      [1]
<a class="moz-txt-link-freetext" href="http://wien2k-algerien1970.blogspot.com/2016/09/the-full-version-of-conversgence-tests.html">http://wien2k-algerien1970.blogspot.com/2016/09/the-full-version-of-conversgence-tests.html</a><br>
      [2] <a class="moz-txt-link-freetext" href="https://www.facebook.com/groups/wien2k.Algeria/">https://www.facebook.com/groups/wien2k.Algeria/</a>
      <br>
      [3]
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15255.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15255.html</a><br>
      [4]
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15258.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg15258.html</a><br>
      [5]
<a class="moz-txt-link-freetext" href="https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19405.html">https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19405.html</a><br>
      [6]
<a class="moz-txt-link-freetext" href="https://www.webassign.net/question_assets/ncsucalcphysmechl3/percent_error/manual.html">https://www.webassign.net/question_assets/ncsucalcphysmechl3/percent_error/manual.html</a><br>
      [7] <a class="moz-txt-link-freetext" href="https://socratic.org/questions/what-percent-error-is-too-high">https://socratic.org/questions/what-percent-error-is-too-high</a>
      <br>
      <br>
      On 12/22/2019 11:36 PM, Peeyush kumar kamlesh wrote:
      <br>
      <blockquote type="cite">Dear Gavin Sir,
        <br>
        Thank you for your response.
        <br>
        I did k-point convergence test by a test package
        "test_conv-5.0.tar.gz". I dont know whether it is gives
        completely satisfactory results or not. If you have any other
        package for kpt convergence kindly let me know, so that i can do
        the right calculations. Also I used non shifted k-mash in the
        calculations of the band gap.
        <br>
        <br>
        Thank you so much!
        <br>
        Regards
        <br>
        Peeyush Kamlesh
        <br>
      </blockquote>
    </blockquote>
  </body>
</html>