<div dir="ltr"><div><div><div>Hi, John, <br><br></div>I have run the code and do not understand how it can be used to reduce the number of constraints in my problem. <br><br></div>Any guidance would be appreciated. <br><br></div>thanks</div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Aug 31, 2016 at 1:29 PM, John Forrest <span dir="ltr"><<a href="mailto:john.forrest@fastercoin.com" target="_blank">john.forrest@fastercoin.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
  
    
  
  <div bgcolor="#FFFFFF" text="#000000">
    <div>Look at Clp/examples/dualCuts.cpp - you
      should be able to adapt it.<br>
      <br>
      John Forrest<div><div class="h5"><br>
      <br>
      On 31/08/16 17:10, usa usa wrote:<br>
    </div></div></div>
    <blockquote type="cite"><div><div class="h5">
      
      <div dir="ltr">
        <p>Hi, </p>
        <p> </p>
        <p>I need to build a
          large scale LP model and solve it by CLP.</p>
        <p> </p>
        <p>In the model, there
          is a kind of constraint like: </p>
        <p><span>     </span></p>
        <p>Max: sum of
          (constantValueP_i<span>  </span>* decVarX_i) from i=1
          to<span>  </span>N</p>
        <p>s.t. </p>
        <p><span>      </span>decVarT + sum of (decVarK_i ) from i=1 to
          I = N<span>  </span><=<span>  </span>[sum of
          (constantValueP_i<span>  </span>* decVarX_i)<span> 
          </span>from i=1 to<span>  </span>N ] * constantQ</p>
        <p><span><br>
          </span></p>
        <p><span>      </span>[sum of (constantValueE_i<span>  </span>*
          decVarX_i)<span> 
          </span>from i=1 to<span>  </span>N ] <= [sum of
          (constantValueE_i )<span>  </span>from i=1 to<span>  </span>N
          ] * constantD</p>
        <p><span><br>
          </span></p>
        <p><span>      </span>decVarK_1 >= sum of
          (constantValue_1_i<span>  </span>* decVarX_i) from i=1
          to<span>  </span>N - <br>
        </p>
        <p>decVarT</p>
        <p><span><br>
          </span></p>
        <p><span>      </span>decVarK_2 >= sum of
          (constantValue_2_i<span>  </span>* decVarX_i) from i=1
          to<span>  </span>N - decVarT</p>
        <p><span><br>
          </span></p>
        <p><span>      </span>…</p>
        <p><span><br>
          </span></p>
        <p><span>      </span>decVarK_L >= sum of
          (constantValue_j_i<span>  </span>* decVarX_i) from i=1
          to<span>  </span>N - decVarT</p>
        <p> </p>
        <p>Decision variables: </p>
        <p>decVarT
          , 0 <= decVarX_i <= 1, decVarK_i >= 0</p>
        <p> </p>
        <p>The problem is that
          the number of constraints of<span>   </span>decVarK_i
          for i=1 to L and L can be very large, e.g. 100,0000. </p>
        <p> </p>
        <p>It means that it
          will have 100,000 constraints in the LP, which I want to
          avoid. </p>
        <p> </p>
        <p>How to combine them
          so that I can reduce the size of the LP model meanwhile
          keeping all constraints
          satisfied ? </p>
        <p> </p>
        <p>thanks</p>
      </div>
      <br>
      <fieldset></fieldset>
      <br>
      </div></div><pre>______________________________<wbr>_________________
Clp mailing list
<a href="mailto:Clp@list.coin-or.org" target="_blank">Clp@list.coin-or.org</a>
<a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__list.coin-2Dor.org_mailman_listinfo_clp&d=CwICAg&c=Ngd-ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLxWPA_2Wlc4&r=js2M0T-3OIMIVDvokcKjokJbk0F8QOCd0mT4FsVFE88&m=S3UeV4Wu5yHELjmWBTDl-eRIeWOSeDtcdfAfTnhn-HU&s=tU_XGRYlmWymVflxTZzAzfVHknXPfwEvd-Sv1KZv2rc&e=" target="_blank">https://urldefense.proofpoint.<wbr>com/v2/url?u=http-3A__list.<wbr>coin-2Dor.org_mailman_<wbr>listinfo_clp&d=CwICAg&c=Ngd-<wbr>ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLx<wbr>WPA_2Wlc4&r=js2M0T-<wbr>3OIMIVDvokcKjokJbk0F8QOCd0mT4F<wbr>sVFE88&m=S3UeV4Wu5yHELjmWBTDl-<wbr>eRIeWOSeDtcdfAfTnhn-HU&s=tU_<wbr>XGRYlmWymVflxTZzAzfVHknXPfwEvd<wbr>-Sv1KZv2rc&e=</a> 
</pre>
    </blockquote>
    <p><br>
    </p>
  </div>

<br>______________________________<wbr>_________________<br>
Clp mailing list<br>
<a href="mailto:Clp@list.coin-or.org">Clp@list.coin-or.org</a><br>
<a href="https://urldefense.proofpoint.com/v2/url?u=http-3A__list.coin-2Dor.org_mailman_listinfo_clp&d=CwICAg&c=Ngd-ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLxWPA_2Wlc4&r=dsAhYgYMFA3lB0slzjj1tRzNWYmsEVkEtULxrYaBfxE&m=OZ7lNGHhz6hncr0dJo35lLoUkFeHcQJj8l5Yd4riBoE&s=QuGUkAd8wEzd7ttcfSt6F7QUcUo8o9NtPe9DXXByuc4&e=" rel="noreferrer" target="_blank">https://urldefense.proofpoint.<wbr>com/v2/url?u=http-3A__list.<wbr>coin-2Dor.org_mailman_<wbr>listinfo_clp&d=CwICAg&c=Ngd-<wbr>ta5yRYsqeUsEDgxhcqsYYY1Xs5ogLx<wbr>WPA_2Wlc4&r=<wbr>dsAhYgYMFA3lB0slzjj1tRzNWYmsEV<wbr>kEtULxrYaBfxE&m=<wbr>OZ7lNGHhz6hncr0dJo35lLoUkFeHcQ<wbr>Jj8l5Yd4riBoE&s=<wbr>QuGUkAd8wEzd7ttcfSt6F7QUcUo8o9<wbr>NtPe9DXXByuc4&e=</a><br>
<br></blockquote></div><br></div>