MINLP with Couenne or Bonmin

[Jan Stampfli]

Hello

I have installed the COIN-OR suite because I want to optimize a MINLP problem. Thereby I tried to use the solver on my computer and on the NEOS server as well.

I have constructed an example  which I wanted to test both solvers:

from __future__ import division                                

from pyomo.environ import*

model = ConcreteModel()

model.x = Var(within=NonNegativeIntegers)

model.y = Var(within=NonNegativeReals)

model.maximizeZ = Objective(expr=model.x,sense=maximize)

model.Constraint1 = Constraint(expr=model.x*model.y<=100)

model.Constraint2 = Constraint(expr=model.x**model.y==20) 

If I try to run it on my computer I get the following errors:

Couenne: 

ERROR: "[base]/site-packages/pyomo/opt/base/solvers.py", 616, solve
        Solver (asl) returned non-zero return code (-6)
ERROR: "[base]/site-packages/pyomo/opt/base/solvers.py", 621, solve
        Solver log:
        Couenne 0.5 -- an Open-Source solver for Mixed Integer Nonlinear Optimization
        Mailing list: couenne@list.coin-or.org
        Instructions: http://www.coin-or.org/Couenne
        couenne:
        ANALYSIS TEST: Exception of type: OPTION_INVALID in file "../../../../Ipopt/src/Algorithm/IpAlgBuilder.cpp" at line 271:
         Exception message: Selected linear solver MA27 not available.
        Tried to obtain MA27 from shared library "libhsl.so", but the following error occured:
        libhsl.so: cannot open shared object file: No such file or directory


        EXIT: Invalid option encountered.
        Loaded instance "/tmp/tmph4CTkG.pyomo.nl"
        Constraints:            2
        Variables:              2 (1 integer)
        Auxiliaries:            5 (1 integer)


        Coin0506I Presolve 11 (-2) rows, 4 (-3) columns and 26 (-8) elements
        Clp0006I 0  Obj -2 Primal inf 7.0209757 (2) Dual inf 3.5738309 (1)
        Clp0006I 2  Obj -171
        Clp0000I Optimal - objective value -171
        Clp0032I Optimal objective -171 - 2 iterations time 0.002, Presolve 0.00
        Clp0000I Optimal - objective value -171
        Couenne: new cutoff value -1.7100000000e+02 (0 seconds)
        NLP Heuristic: terminate called after throwing an instance of 'Ipopt::INVALID_WARMSTART'
[    0.34] Pyomo Finished
ERROR: Unexpected exception while running model:
        Solver (asl) did not exit normally

Bonmin:

ERROR: Unexpected exception while running model:
        Cannot load a SolverResults object with bad status: error

If I try to run the solvers on the NEOS server I get the following errors:

Couenne:

ERROR: Error parsing NEOS solution file  NEOS log:
        Job 5741310 dispatched
        password: VNnMQexW
        ---------- Begin Solver Output -----------
        Condor submit: 'neos.submit'
        Condor submit: 'watchdog.submit'
        Job submitted to NEOS HTCondor pool.
ERROR: ERROR: An error occured with your submission.
[   22.25] Pyomo Finished
ERROR: Unexpected exception while running model:
        Problem executing an event.  No results are available.

Bonmin:

ERROR: Error parsing NEOS solution file  NEOS log:
        Job 5741308 dispatched
        password: eZtrAWlu
        ---------- Begin Solver Output -----------
        Condor submit: 'neos.submit'
        Condor submit: 'watchdog.submit'
        Job submitted to NEOS HTCondor pool.
ERROR: ERROR: An error occured with your submission.
[    7.75] Pyomo Finished
ERROR: Unexpected exception while running model:
        Problem executing an event.  No results are available.

In the end, I want to use the NEOS server to solve my problem.

[Jan Stampfli]

Due to the fact, that I could not find out the problem I installed all my stuff now on windows. All solver of the COIN-OR suite work now. But the connection to the NEOS server does not work anymore.  If i try to run for e.g.

pyomo solve --solver=ipopt  --solver-manager=neos rosenbrock.py

i get the error 

pyomo-script.py solve: error: unrecognized arguments: --solver-manager=neos

Does anyone know why pyomo the command --solver-manager not recognizes?

When I edit my script to include the solver-manager as follows

from pyomo.environ import *

model = AbstractModel()

model.x = Var(initialize=1.5)

model.y = Var(initialize=1.5)

def rosenbrock(model):

        return (1.0-model.x)**2 + 100.0*(model.y - model.x**2)**2 

model.obj = Objective(rule=rosenbrock, sense=minimize)

with SolverManagerFactory("neos") as manager:

        manager.solve(model, opt="ipopt")

and then run 

python rosenbrock.py

 

I get the following error:

Job 5743071 submitted to NEOS, password='zvxLrndZ'

Check the following URL for progress report :

https://www.neos-server.org/neos/cgi-bin/nph-neos-solver.cgi?admin=results&jobnumber=5743071&pass=zvxLrndZ

Job 5743071 dispatched

password: zvxLrndZ

---------- Begin Solver Output -----------

Condor submit: 'neos.submit'

Condor submit: 'watchdog.submit'

Job submitted to NEOS HTCondor pool.

Traceback (most recent call last):

  File "rosenbrock.py", line 16, in <module>

    manager.solve(model, opt="ipopt")

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\async_solver.py", line 33, in solve

    return self.execute(*args, **kwds)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\manager.py", line 107, in execute

    results = self.wait_for(ah)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\manager.py", line 170, in wait_for

    tmp = self.wait_any()

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\manager.py", line 159, in wait_any

    ah = self._perform_wait_any()

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\neos\plugins\kestrel_plugin.py", line 185, in _perform_wait_any

    solver_results = opt.process_output(rc)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\solver\shellcmd.py", line 348, in process_output

    suffixes=self._suffixes)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\plugins\sol.py", line 54, in __call__

    % (filename, str(e), fdata))

ValueError: Error reading 'c:\users\janst\appdata\local\temp\tmpcuvsva.neos.sol': no Options line found.

SOL File Output:

ERROR: An error occured with your submission.

I also tried 

from pyomo.environ import *

model = AbstractModel()

model.x = Var(initialize=1.5)

model.y = Var(initialize=1.5)

def rosenbrock(model):

        return (1.0-model.x)**2 + 100.0*(model.y - model.x**2)**2 

model.obj = Objective(rule=rosenbrock, sense=minimize)

#with SolverManagerFactory("neos") as manager:

#        manager.solve(model, opt="ipopt")

opt = SolverFactory('ipopt')

solver_manager = SolverManagerFactory('neos')

results = solver_manager.solve(model, opt=opt)

results.write()

which gives me the following error:

Traceback (most recent call last):

  File "rosenbrock.py", line 20, in <module>

    results = solver_manager.solve(model, opt=opt)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\async_solver.py", line 33, in solve

    return self.execute(*args, **kwds)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\manager.py", line 106, in execute

    ah = self.queue(*args, **kwds)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\opt\parallel\manager.py", line 121, in queue

    return self._perform_queue(ah, *args, **kwds)

  File "C:\Users\janst\AppData\Roaming\Python\Python27\site-packages\pyomo\neos\plugins\kestrel_plugin.py", line 96, in _perform_queue

    % (solver_name, str(sorted(self._solvers.keys()))))

pyomo.opt.parallel.manager.ActionManagerError: Solver 'couenne.exe' is not recognized by NEOS. Solver names recognized:

['bonmin', 'cbc', 'conopt', 'couenne', 'cplex', 'filmint', 'filter', 'ipopt', 'knitro', 'l-bfgs-b', 'lancelot', 'loqo', 'minlp', 'minos', 'minto', 'mosek', 'ooqp', 'path', 'pcx', 'snopt']

[fant...@gmail.com]

Hi Jan,

May I please ask have you solved this issue? I'm solving an MINLP too, with minimization of a log-sum-exp example (using Baron on Pyomo though), and I run into similar error. So I'm just wondering if you have found a solution to it. 

Thanks!

[Jan Stampfli]

Hi fantian93

I could not solve this issue with NEOS. I had the opportunity to solve my optimization on a server from my university.

Regards

Jan

--
You received this message because you are subscribed to a topic in the Google Groups "Pyomo Forum" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pyomo-forum/mzTWTnqQ_HQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pyomo-forum+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.