<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
I've seen this as well on all kinds of problems. Our problems have
a lot of equality constraints, I found by accident that setting
"bound_relax_factor" to 0.0 to help with equalities also greatly
reduced changes from run to run. I never chased up why.<br>
<br>
Damien<br>
<br>
<div class="moz-cite-prefix">On 2014-11-13 4:45 PM, Tony Kelman
wrote:<br>
</div>
<blockquote cite="mid:1603FCA3C83B49A4A96482539F72511D@TKsamsung"
type="cite">
<div dir="ltr">
<div style="FONT-SIZE: 12pt; FONT-FAMILY: 'Calibri'; COLOR:
#000000">
<div>The most common case of non-determinism is multithreaded
BLAS/LAPACK libraries from MKL or OpenBLAS. Those are
generally not exactly deterministic beyond machine precision
unless you run them with a single thread. Some of the
multithreaded sparse linear solvers like MA86 are also
non-deterministic due to parallelism. If you really want to
hunt this down, as I said in those linked threads from last
year, set a high print level and output the logs to
different files, to compare whether everything is truly
starting at exactly the same point, and if so, where things
start differing and by how much. With complicated non-convex
problems, it’s entirely possible that initially very minor
numerical roundoff differences can accumulate and eventually
lead Ipopt down a different convergence path to a different
solution.</div>
<div> </div>
<div>-Tony</div>
<div> </div>
<div style="FONT-SIZE: small; TEXT-DECORATION: none;
FONT-FAMILY: "Calibri"; FONT-WEIGHT: normal;
COLOR: #000000; FONT-STYLE: normal; DISPLAY: inline">
<div style="FONT: 10pt tahoma">
<div> </div>
<div style="BACKGROUND: #f5f5f5">
<div style="font-color: black"><b>From:</b> <a
moz-do-not-send="true" title="ipoptuser@gmail.com"
href="mailto:ipoptuser@gmail.com">Ipopt User</a> </div>
<div><b>Sent:</b> Thursday, November 13, 2014 5:35 AM</div>
<div><b>To:</b> <a moz-do-not-send="true"
title="Guy.Charles@redbulltechnology.com"
href="mailto:Guy.Charles@redbulltechnology.com">Guy
Charles</a> </div>
<div><b>Cc:</b> <a moz-do-not-send="true"
title="ipopt@list.coin-or.org"
href="mailto:ipopt@list.coin-or.org">ipopt@list.coin-or.org</a>
</div>
<div><b>Subject:</b> Re: [Ipopt] Is ipopt deterministic?</div>
</div>
</div>
<div> </div>
</div>
<div style="FONT-SIZE: small; TEXT-DECORATION: none;
FONT-FAMILY: "Calibri"; FONT-WEIGHT: normal;
COLOR: #000000; FONT-STYLE: normal; DISPLAY: inline">
<div dir="ltr">IPOPT itself is fully deterministic, but some
dependencies may not be deterministic. Please see <a
moz-do-not-send="true"
href="http://list.coin-or.org/pipermail/ipopt/2013-April/003337.html">http://list.coin-or.org/pipermail/ipopt/2013-April/003337.html</a>
and <a moz-do-not-send="true"
href="http://list.coin-or.org/pipermail/ipopt-tickets/2013-March/000635.html">http://list.coin-or.org/pipermail/ipopt-tickets/2013-March/000635.html</a>
for possible explanations of the behavior you are seeing.
</div>
<div class="gmail_extra">
<div> </div>
<div class="gmail_quote">On Thu, Nov 13, 2014 at 6:35 AM,
Guy Charles <span dir="ltr"><<a
moz-do-not-send="true"
href="mailto:Guy.Charles@redbulltechnology.com"
target="_blank">Guy.Charles@redbulltechnology.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="PADDING-LEFT:
1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px
solid">
<div vlink="#954F72" link="#0563C1" lang="EN-GB">
<div>
<p class="MsoNormal"><span style="COLOR: #1f497d">Hi,</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d">More
out of curiosity really: </span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d">At
times I’m getting some non-deterministic
behaviour in my iterations and no of
iterations before Ipopt finds an optimum.
Despite running the same script twice in a row
with identical conditions, I may get different
results. I assume that the ‘error’ is in my
function calls, scripts or layers of code
between ipopt and the function calls, but I
thought I should double check that Ipopt is
completely deterministic.</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Should
ipopt be completely deterministic from each
isolated solve call?</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Is
there any random jitter added in when it
encounters an issue? </span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Would
a badly scaled problem cause non-deterministic
issues w.r.t numerical precision? </span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Is
it possibly that anything in the linear
solvers that could cause this?</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Thanks
in advance for any help,</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d">Guy</span></p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
<p class="MsoNormal"><span style="COLOR: #1f497d"></span> </p>
</div>
<br clear="all">
______________________________________________________________________<br>
This email has been scanned by the Symantec Email
Security.cloud service.<br>
For more information please visit <a
moz-do-not-send="true"
href="http://www.symanteccloud.com"
target="_blank">http://www.symanteccloud.com</a><br>
______________________________________________________________________<br>
</div>
<br>
_______________________________________________<br>
Ipopt mailing list<br>
<a moz-do-not-send="true"
href="mailto:Ipopt@list.coin-or.org">Ipopt@list.coin-or.org</a><br>
<a moz-do-not-send="true"
href="http://list.coin-or.org/mailman/listinfo/ipopt"
target="_blank">http://list.coin-or.org/mailman/listinfo/ipopt</a><br>
<br>
</blockquote>
</div>
<div> </div>
</div>
<p>
</p>
<hr>
_______________________________________________<br>
Ipopt mailing list<br>
<a class="moz-txt-link-abbreviated" href="mailto:Ipopt@list.coin-or.org">Ipopt@list.coin-or.org</a><br>
<a class="moz-txt-link-freetext" href="http://list.coin-or.org/mailman/listinfo/ipopt">http://list.coin-or.org/mailman/listinfo/ipopt</a><br>
</div>
</div>
</div>
<br>
<fieldset class="mimeAttachmentHeader"></fieldset>
<br>
<pre wrap="">_______________________________________________
Ipopt mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Ipopt@list.coin-or.org">Ipopt@list.coin-or.org</a>
<a class="moz-txt-link-freetext" href="http://list.coin-or.org/mailman/listinfo/ipopt">http://list.coin-or.org/mailman/listinfo/ipopt</a>
</pre>
</blockquote>
<br>
</body>
</html>