[Ipopt] some iterations taking much longer than others

[david greer]

Hi,

for the problem I am currently working on, I have noticed that some iterations only take 2 or 3 seconds, while others take much longer, around 20 seconds.  I was wondering what might be causing this large variability.  The only explanation I can think of is that some iterations require many trials when perturbing the Hessian to achieve the expected matrix inertia and so are performing many matrix decompositions.  Is that a plausible explanation?  If so, is there some way I can restructure my problem to reduce the likelihood of time-consuming iterations?  If not, can anyone suggest a better explanation?

Best wishes,

David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.coin-or.org/pipermail/ipopt/attachments/20140306/9d312d5f/attachment.html>