[Ipopt] IpOpt code: Detereministic or Random across various runs?
AliReza Khoshgoftar Monfared
khoshgoftar at gmail.com
Mon Dec 17 21:41:34 EST 2012
Hi giacomo,
The problems are exactly the same across various runs. What i was
referring by similar was succession of problems solved in a single
run. Despite this each run solves exactly the same problem with the
same initial point.
Alireza
Sent from my iPhone
On Dec 17, 2012, at 7:48 PM, Giacomo Nannicini <giacomo.n at gmail.com> wrote:
> Hello,
> are the problems you are trying to solve identical or just "very
> similar" as you state below?
> If they are "very similar" but not identical, it's no surprise that
> you get a variation in the order of 1e-3 at some point during the
> algorithm, and then this difference can increase to larger values as
> the two runs start to diverge.
> Floating point computations are very sensitive to the input data.
>
> If your problems are absolutely identical and you provide exactly the
> same starting point, then it's worth investigating where the two runs
> start to diverge.
>
> Giacomo
>
> On Tue, Dec 18, 2012 at 4:28 AM, AliReza Khoshgoftar Monfared
> <khoshgoftar at gmail.com> wrote:
>> Thanks Stefan,
>>
>> I tested the error proposed in the ticket and since I could reproduce it, I
>> applied the patch. Now it's gone.
>> Despite this, I still see the deviation in answers to the problem I have.
>> To clear the way this error is produced, I solve three very similar
>> optimization problems in succession (say t=0, t=1 and t=2 problems). I run
>> this two times, and for each run, I provide Ipopt with the exact answer for
>> t=0. Since problems at t=1 and 2 are similar, it is expected that answers
>> are the same, but I see that at t=1 there is a deviation on the order of
>> 1e-4 across the 2 runs and at t=2 it's on the order of 1e-3 (and this error
>> can propagate larger if i continue this)
>> As stated, despite this, Ipopt is giving me answers both of which respect
>> all (non-linear) constraints of the problem, so they are all correct answers
>> in this sense, although I expect ipopt to follow a determinism path and
>> arrive at the same correct answer from the same initial point every time.
>>
>> I will try to find the differences in the ipopt report of problems and post
>> them here if I could figure out the reason.
>>
>> Alireza
>>
>>
More information about the Ipopt
mailing list