<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <font face="monospace">Dear Alexander,<br>
      <br>
      I assume that in your problem description you are using the LP
      format. However, the documentation<br>
      <br>
<a class="moz-txt-link-freetext" href="https://www.tu-chemnitz.de/mathematik/discrete/manuals/cplex/doc/refman/html/appendixE5.html">https://www.tu-chemnitz.de/mathematik/discrete/manuals/cplex/doc/refman/html/appendixE5.html</a><br>
      <br>
      states "... <, <=, =<, >, >=, =>, and =. These
      are interpreted as <=, <=, <=, >=, >=, >= and =,
      respectively". The reason is that linear programs (which are used
      as relaxations for IPs) in general do not allow strict
      inequalities.<br>
      <br>
      Best,<br>
      Matthias<br>
    </font><br>
    <div class="moz-cite-prefix">On 03.03.22 01:17, Alex Meiburg wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CACs9yV81ZhTEKexVEfiZsNwb9txHXqM_THtXGwO=Sjs52SwXWg@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="ltr">
        <div>I noticed that SCIP doesn't seem to like strict
          inequalities for integers. When it solves</div>
        <div><br>
        </div>
        <div>Maximize<br>
          k<br>
          st<br>
          k < 10<br>
          Integer<br>
          k</div>
        <div><br>
        </div>
        <div>I get the solution</div>
        <div><br>
        </div>
        <div>solution status: optimal solution found<br>
          objective value:                                   10<br>
          k                                                  10 (obj:1)</div>
        <div><br>
        </div>
        <div>When of course that's not actually feasible. :) In this
          case of course it's pretty simple to remedy by replacing
          "[foo] < N" by "[foo] <= N-1". But that's not always
          valid, e.g. in the problem</div>
        <div><br>
        </div>
        <div>Maximize<br>
          halfK<br>
          st<br>
        </div>
        <div>halfK < 10</div>
        <div>halfK + halfK = k<br>
        </div>
        <div>Integer<br>
          k</div>
        <div><br>
        </div>
        <div>... in which case the correct answer is 9.5 (and SCIP think
          it's 10). I'm using SCIP v7.0.2. Is there any interest in
          fixing this, or at least, documenting it better...? I skimmed
          and couldn't find any warnings about not using strict
          inequalities.<br>
        </div>
        <div><br>
        </div>
        <div>
          <div>
            <div dir="ltr" class="gmail_signature"
              data-smartmail="gmail_signature">
              <div dir="ltr">-- Alexander Meiburg<br>
              </div>
            </div>
          </div>
        </div>
      </div>
      <br>
      <fieldset class="moz-mime-attachment-header"></fieldset>
      <pre class="moz-quote-pre" wrap="">_______________________________________________
Scip mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Scip@zib.de">Scip@zib.de</a>
<a class="moz-txt-link-freetext" href="https://listserv.zib.de/mailman/listinfo/scip">https://listserv.zib.de/mailman/listinfo/scip</a>
</pre>
    </blockquote>
    <br>
  </body>
</html>