commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexey Slepov (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MATH-828) Not expected UnboundedSolutionException
Date Mon, 23 Jul 2012 10:03:35 GMT

    [ https://issues.apache.org/jira/browse/MATH-828?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13420534#comment-13420534
] 

Alexey Slepov commented on MATH-828:
------------------------------------

Now with 3.1-SNAPSHOT things are going better. Number of unbounded exceptions and unexpected
theta values (> 1) are fewer versus math 3.0.
Now the picture of failures and successes for
25 entities (i.e. 25 + 1 variables) with 7 constraint equations, 3 for -q(r) + QL >= 0
and 4 for x(r)*t - XL >= 0
looks like

Iteration 1 of 64
Iteration 2 of 64
Iteration 3 of 64
Iteration 4 of 64
Iteration 5 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 6 of 64
Iteration 7 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 8 of 64
Iteration 9 of 64
Iteration 10 of 64
Iteration 11 of 64
Iteration 12 of 64
Iteration 13 of 64
Iteration 14 of 64
Iteration 15 of 64
Iteration 16 of 64
Iteration 17 of 64
Iteration 18 of 64
Iteration 19 of 64
Iteration 20 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 21 of 64
Iteration 22 of 64
Iteration 23 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 24 of 64
Iteration 25 of 64
Iteration 26 of 64
EXCEPTION: unbounded solution
Iteration 27 of 64
Iteration 28 of 64
Iteration 29 of 64
Iteration 30 of 64
Iteration 31 of 64
Iteration 32 of 64
Iteration 33 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 34 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 35 of 64
Iteration 36 of 64
Iteration 37 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 38 of 64
Iteration 39 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 40 of 64
Iteration 41 of 64
Iteration 42 of 64
Iteration 43 of 64
Iteration 44 of 64
Iteration 45 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 46 of 64
Iteration 47 of 64
Iteration 48 of 64
Iteration 49 of 64
Iteration 50 of 64
Iteration 51 of 64
Iteration 52 of 64
Iteration 53 of 64
Iteration 54 of 64
Iteration 55 of 64
Iteration 56 of 64
Iteration 57 of 64
Iteration 58 of 64
Iteration 59 of 64
Iteration 60 of 64
Iteration 61 of 64
Iteration 62 of 64
EXCEPTION: illegal state: maximal count (32,768) exceeded
Iteration 63 of 64
Iteration 64 of 64

for the code
SimplexSolver solver = new SimplexSolver(epsilon, 15);
try
{
	solver.setMaxIterations(32768);
	PointValuePair optimum = solver.optimize(objectiveFunction, constraints, GoalType.MINIMIZE,
true);
...

It's much better but it's still to risky to use math 3 to solve problems of this kind in real-world
projects.
                
> Not expected UnboundedSolutionException
> ---------------------------------------
>
>                 Key: MATH-828
>                 URL: https://issues.apache.org/jira/browse/MATH-828
>             Project: Commons Math
>          Issue Type: Bug
>    Affects Versions: 3.0
>         Environment: Intel Core i5-2300 Windows XP SP3
>            Reporter: Alexey Slepov
>              Labels: linear, math, programming
>             Fix For: 3.1
>
>         Attachments: ApacheSimplexWrapper.java, ApacheSimplexWrapperTest.java, Entity.java,
commons-math3-3.0.jar
>
>
> SimplexSolver throws UnboundedSolutionException when trying to solve minimization linear
programming problem. The number of exception thrown depends on the number of variables.
> In order to see that behavior of SimplexSolver first try to run JUnit test setting a
final variable ENTITIES_COUNT = 2 and that will give almost good result and then set it to
15 and you'll get a massive of unbounded exceptions.
> First iteration is runned with predefined set of input data with which the Solver gives
back an appropriate result.
> The problem itself is well tested by it's authors (mathematicians who I believe know
what they developed) using Matlab 10 with no unbounded solutions on the same rules of creatnig
random variables values.
> What is strange to me is the dependence of the number of UnboundedSolutionException exceptions
on the number of variables in the problem.
> The problem is formulated as
> min(1*t + 0*L) (for every r-th subject)
> s.t.
> -q(r) + QL >= 0
> x(r)t - XL >= 0
> L >= 0
> where 
> r = 1..R, 
> L = {l(1), l(2), ..., l(R)} (vector of R rows and 1 column),
> Q - coefficients matrix MxR
> X - coefficients matrix NxR 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message