Maximum feasibility tolerance

GAMS
1

Hi,
I’m working with a nlp model and the following message appears to
me:

Infeasible solution. A free variable exceeds the allowable range.
Current value is 3.37E+10 and current upper bound (Rtmaxv) is 1.00E
+10

how I could change the Maximum feasibility tolerance?

thank you

\

2

Arne,

I reduced the monetary values ​​for billions of $ and other variables are rates. The message that now appears is as follows:
** Infeasible solution. Reduced gradient less than tolerance.
you have any suggestions for me to identify the problem?


Regards,
Cláudia



MODEL STATISTICS
BLOCKS OF EQUATIONS 66 SINGLE EQUATIONS 2,016
BLOCKS OF VARIABLES 53 SINGLE VARIABLES 2,011
NON ZERO ELEMENTS 9,791 NON LINEAR N-Z 3,798
DERIVATIVE POOL 6 CONSTANT POOL 197
CODE LENGTH 11,844

GENERATION TIME = 0.343 SECONDS 5 Mb WIN236-236 Mar 30, 2011

EXECUTION TIME = 0.359 SECONDS 5 Mb WIN236-236 Mar 30, 2011
Solution Report SOLVE csonaglio Using NLP From line 1296

S O L V E S U M M A R Y
MODEL csonaglio OBJECTIVE GDPSUM
TYPE NLP DIRECTION MINIMIZE
SOLVER CONOPT FROM LINE 1296
**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 5 Locally Infeasible
**** OBJECTIVE VALUE 0.0000
RESOURCE USAGE, LIMIT 136.252 10000.000
ITERATION COUNT, LIMIT 1798 9000
EVALUATION ERRORS 0 0
CONOPTD 0.1 Dec 13, 2010 23.6.4 WIN 23843.23989 VS8 x86/MS Windows


C O N O P T 3 version 3.14W
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark


** Infeasible solution. Reduced gradient less than tolerance.







On Fri, Oct 28, 2011 at 6:44 AM, Arne Stolbjerg Drud wrote:

Claudia:

This is not a problem with the feasibility tolerance. You have very large
variables. You can scale the part of the model with these large variables
(this mean that you measure them in a different unit so the numerical value
becomes smaller). An alternative is to define the option Rtmaxv (mentioned
in the message) to a larger value than the default, for example Rtmaxv =
1.e15. This is done in the conopt.opt option file and you must remember to
tell the solver to use an options file in the usual way.

Good luck

Arne

Arne Stolbjerg Drud
ARKI Consulting & Development A/S
Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark
Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: adrud@arki.dk

-----Oprindelig meddelelse-----
Fra: gamsworld@googlegroups.com [mailto:gamsworld@googlegroups.com] PÃ¥ vegne
af Claudia Sonaglio
Sendt: 27. oktober 2011 21:53
Til: gamsworld
Emne: Maximum feasibility tolerance


Hi,
I’m working with a nlp model and the following message appears to
me:

Infeasible solution. A free variable exceeds the allowable range.
Current value is 3.37E+10 and current upper bound (Rtmaxv) is 1.00E
+10

how I could change the Maximum feasibility tolerance?

thank you

–
“gamsworld” group.
To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to
gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/gamsworld?hl=en.

\

To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

\

To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

3

Arne,

The optimized model (great place), but returned the following message during interactions.

**** Non-Fatal Internal Error Number 1011
The solution process will recover and continue correctly.

Please check that you have the latest version of CONOPT and
upgrade your system – the problem may already be fixed.
If your system is up to date, please send a copy of your
model (input and output files) by e-mail to adrud@arki.dk


** Optimal solution. Reduced gradient less than tolerance

I’m using version 23.6. Any ideas for this situation?


Regards,
Cláudia

S O L V E S U M M A R Y
MODEL csonaglio OBJECTIVE GDPSUM
TYPE NLP DIRECTION MINIMIZE
SOLVER CONOPT FROM LINE 1285
**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 2 Locally Optimal
**** OBJECTIVE VALUE 0.0000
RESOURCE USAGE, LIMIT 154.924 10000.000
ITERATION COUNT, LIMIT 857 9000
EVALUATION ERRORS 0 0
CONOPTD 0.1 Dec 13, 2010 23.6.4 WIN 23843.23989 VS8 x86/MS Windows


C O N O P T 3 version 3.14W
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark



On Mon, Oct 31, 2011 at 5:06 PM, Arne Stolbjerg Drud wrote:

Claudia,



Your model could be infeasible – difficult to tell. At least CONOPT did a lot of work before stopping here. If you are sure there is a solution then you should look at the solution – if some variables are far away from where you expect the solution should be then you should give these variables some initial values that are closer to where you think it should be.



I am afraid there is not much more I can do / suggest based on this information.



Regards



Arne


\

Arne Stolbjerg Drud

ARKI Consulting & Development A/S

Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark

Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: adrud@arki.dk



Fra: gamsworld@googlegroups.com [mailto:gamsworld@googlegroups.com] PÃ¥ vegne af Claudia Sonaglio
Sendt: 31. oktober 2011 18:57
Til: gamsworld@googlegroups.com
Emne: Re: Maximum feasibility tolerance



Arne,



I reduced the monetary values ​​for billions of $ and other variables are rates. The message that now appears is as follows:

** Infeasible solution. Reduced gradient less than tolerance.

you have any suggestions for me to identify the problem?





Regards,

Cláudia







MODEL STATISTICS

BLOCKS OF EQUATIONS 66 SINGLE EQUATIONS 2,016
BLOCKS OF VARIABLES 53 SINGLE VARIABLES 2,011
NON ZERO ELEMENTS 9,791 NON LINEAR N-Z 3,798
DERIVATIVE POOL 6 CONSTANT POOL 197
CODE LENGTH 11,844


GENERATION TIME = 0.343 SECONDS 5 Mb WIN236-236 Mar 30, 2011


EXECUTION TIME = 0.359 SECONDS 5 Mb WIN236-236 Mar 30, 2011

Solution Report SOLVE csonaglio Using NLP From line 1296


S O L V E S U M M A R Y

MODEL csonaglio OBJECTIVE GDPSUM
TYPE NLP DIRECTION MINIMIZE
SOLVER CONOPT FROM LINE 1296

**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 5 Locally Infeasible
**** OBJECTIVE VALUE 0.0000

RESOURCE USAGE, LIMIT 136.252 10000.000
ITERATION COUNT, LIMIT 1798 9000
EVALUATION ERRORS 0 0
CONOPTD 0.1 Dec 13, 2010 23.6.4 WIN 23843.23989 VS8 x86/MS Windows


C O N O P T 3 version 3.14W
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark


** Infeasible solution. Reduced gradient less than tolerance.











On Fri, Oct 28, 2011 at 6:44 AM, Arne Stolbjerg Drud wrote:

Claudia:

This is not a problem with the feasibility tolerance. You have very large
variables. You can scale the part of the model with these large variables
(this mean that you measure them in a different unit so the numerical value
becomes smaller). An alternative is to define the option Rtmaxv (mentioned
in the message) to a larger value than the default, for example Rtmaxv =
1.e15. This is done in the conopt.opt option file and you must remember to
tell the solver to use an options file in the usual way.

Good luck

Arne

Arne Stolbjerg Drud
ARKI Consulting & Development A/S
Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark
Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: adrud@arki.dk

-----Oprindelig meddelelse-----
Fra: gamsworld@googlegroups.com [mailto:gamsworld@googlegroups.com] PÃ¥ vegne
af Claudia Sonaglio
Sendt: 27. oktober 2011 21:53
Til: gamsworld
Emne: Maximum feasibility tolerance



Hi,
I’m working with a nlp model and the following message appears to
me:

Infeasible solution. A free variable exceeds the allowable range.
Current value is 3.37E+10 and current upper bound (Rtmaxv) is 1.00E
+10

how I could change the Maximum feasibility tolerance?

thank you

–
“gamsworld” group.
To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to
gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/gamsworld?hl=en.

\

To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.


\

To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

–
To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

\

To post to this group, send email to gamsworld@googlegroups.com.
To unsubscribe from this group, send email to gamsworld+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gamsworld?hl=en.

4

Reply-to: gamsworld@googlegroups.com

Dear All,

I hope I find you well. Please assist me. I have run a model and found the following results below. I have noticed that i cannot get the detailed results like those shown in http://www.gams.com/dd/docs/gams/Tutorial.pdf on page 24. Are these results OK given the model status is 2 Locally Optimal. Also what do I do in order to get results for the other equations in the model, now I just see the final objective value.

Thank you,

Kind regards,

Loveness

Solution Report SOLVE MODEL2 Using NLP From line 1110


S O L V E S U M M A R Y

MODEL MODEL2 OBJECTIVE OBJ
TYPE NLP DIRECTION MAXIMIZE
SOLVER CONOPT3 FROM LINE 1110

**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 2 Locally Optimal
**** OBJECTIVE VALUE 4648075562.5682

RESOURCE USAGE, LIMIT 0.000 50000.000
ITERATION COUNT, LIMIT 6 50000
EVALUATION ERRORS 0 0
CONOPT 3 24.1.3 r41464 Released Jul 26, 2013 VS8 x86/MS Windows


C O N O P T 3 version 3.15L
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark


The model has 315 variables and 224 constraints
with 426 Jacobian elements, 30 of which are nonlinear.
The Hessian of the Lagrangian has 18 elements on the diagonal,
12 elements below the diagonal, and 30 nonlinear variables.

** Optimal solution. Reduced gradient less than tolerance.


CONOPT time Total 0.008 seconds
of which: Function evaluations 0.001 = 12.5%
1st Derivative evaluations 0.000 = 0.0%



**** REPORT SUMMARY : 0 NONOPT
0 INFEASIBLE
0 UNBOUNDED
0 ERRORS
GAMS 24.1.3 r41464 Released Jul 26, 2013 WEX-VS8 x86/MS Windows 08/12/15 01:28:41 Page 6
G e n e r a l A l g e b r a i c M o d e l i n g S y s t e m
Solution Report SOLVE MODEL2 Using NLP From line 1110


**** SOLVER STATUS FILE LISTED BELOW
=CCONOPT 3 24.1.3 r41464 Released Jul 26, 2013 VS8 x86/MS Windows
=0CONOPT 3 24.1.3 r41464 Released Jul 26, 2013 VS8 x86/MS Windows
=1


C O N O P T 3 version 3.15L
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark

=2
1 INPUT (D
EFAULT)
2 BRANCH * MAJERR MINERR JACERR (D
EFAULT)
3 OPTIMIZE (D
EFAULT)
4 BRANCH * * * * * * MAJERR (D
EFAULT)
5 OUTPUT (D
EFAULT)
6 EXIT (D
EFAULT)
7 MAJERR MESSAGE MAJOR ERROR - TERMINATE (D
EFAULT)
8 EXIT (D
EFAULT)
9 MINERR MESSAGE MINOR ERROR - TERMINATE (D
EFAULT)
10 EXIT (D
EFAULT)
11 JACERR MESSAGE FUNCTION OR DERIVATIVE ERROR (D
EFAULT)
12 OUTPUT (D
EFAULT)
13 EXIT (D
EFAULT)

*** Line 1 INPUT

*** Line 0 SET LFEERR = 0

*** Line 0 SET LFITER = 50000

*** Line 0 SET RVTIME = 50000.00000
=1

The model has 315 variables and 224 constraints
with 426 Jacobian elements, 30 of which are nonlinear.
The Hessian of the Lagrangian has 18 elements on the diagonal,
12 elements below the diagonal, and 30 nonlinear variables.
=2

*** Line 2 BRANCH with Lbranch = 1

*** Line 3 OPTIMIZE

Search for a first feasible solution


Iter Phase Ninf Infeasibility RGmax NSB Step InItr MX OK
0 0 7.3390285070E+08 (Input point)
Pre-triangular equations: 92
Post-triangular equations: 105
1 0 1.2678000000E+03 (After pre-processing)
2 0 3.2562109375E+01 (After scaling)
3 0 21 2.5362109375E+01 1.0E+00 F T
4 1 2 1.2633789062E+00 1.0E+00 24 1.0E+00 3 T T

** Feasible solution. Value of objective = 4648075562.50

Search for an optimal solution.


Iter Phase Ninf Objective RGmax NSB Step InItr MX OK
5 3 4.6480755626E+09 9.6E-02 6 1.0E+00 6 F T
6 3 4.6480755626E+09 0.0E+00 6
=1

** Optimal solution. Reduced gradient less than tolerance.

=2

*** Line 4 BRANCH with Lbranch = 1

*** Line 5 OUTPUT

*** Line 6 EXIT
=1

CONOPT time Total 0.008 seconds
of which: Function evaluations 0.001 = 12.5%
1st Derivative evaluations 0.000 = 0.0%

=2

Work length = 100133 double words = 0.76 Mbytes
Estimate = 100133 double words = 0.76 Mbytes
Max used = 32988 double words = 0.25 Mbytes

Timing for Function Evaluations:

Statistics for FDEval-Fnc Calls: 126. Time: 0.001 T/C: 7.9365E-06
Statistics for FDEval-Drv Calls: 90. Time: 0.000 T/C: 0.0000E+00
Statistics for 2DDirLag Calls: 0. Time: 0.000 T/C: 0.0000E+00
Statistics for 2DLagr Calls: 0. Time: 0.000 T/C: 0.0000E+00

**** SOLVER STATUS FILE LISTED ABOVE
GAMS 24.1.3 r41464 Released Jul 26, 2013 WEX-VS8 x86/MS Windows 08/12/15 01:28:41 Page 7
G e n e r a l A l g e b r a i c M o d e l i n g S y s t e m
Solution Report SOLVE MODEL2 Using NLP From line 1110


EXECUTION TIME = 0.031 SECONDS 2 MB 24.1.3 r41464 WEX-VS8




On Tuesday, November 1, 2011 at 5:47:23 PM UTC+1, Arne Stolbjerg Drud wrote:

Claudia



The message means that CONOPT found some internal inconsistency. However, it restarted the optimization and the solution you have is a (local) optimum so you can safely continue with your work.



If you can send me the model (input file(s) + the listing file) I would like to locate and fix this error so an improved version of CONOPT can appear in a future GAMS release.



Regards



Arne


\

Arne Stolbjerg Drud

ARKI Consulting & Development A/S

Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark

Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: ad…@arki.dk



Fra: gams...@googlegroups.com [mailto:gams...@googlegroups.com] PÃ¥ vegne af Claudia Sonaglio
Sendt: 1. november 2011 16:10
Til: gams...@googlegroups.com
Emne: Re: Maximum feasibility tolerance



Arne,



The optimized model (great place), but returned the following message during interactions.



**** Non-Fatal Internal Error Number 1011
The solution process will recover and continue correctly.

Please check that you have the latest version of CONOPT and
upgrade your system – the problem may already be fixed.
If your system is up to date, please send a copy of your
model (input and output files) by e-mail to ad…@arki.dk


** Optimal solution. Reduced gradient less than tolerance



I’m using version 23.6. Any ideas for this situation?





Regards,

Cláudia



S O L V E S U M M A R Y

MODEL csonaglio OBJECTIVE GDPSUM
TYPE NLP DIRECTION MINIMIZE
SOLVER CONOPT FROM LINE 1285

**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 2 Locally Optimal
**** OBJECTIVE VALUE 0.0000

RESOURCE USAGE, LIMIT 154.924 10000.000
ITERATION COUNT, LIMIT 857 9000
EVALUATION ERRORS 0 0
CONOPTD 0.1 Dec 13, 2010 23.6.4 WIN 23843.23989 VS8 x86/MS Windows


C O N O P T 3 version 3.14W
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark



On Mon, Oct 31, 2011 at 5:06 PM, Arne Stolbjerg Drud wrote:

Claudia,



Your model could be infeasible – difficult to tell. At least CONOPT did a lot of work before stopping here. If you are sure there is a solution then you should look at the solution – if some variables are far away from where you expect the solution should be then you should give these variables some initial values that are closer to where you think it should be.



I am afraid there is not much more I can do / suggest based on this information.



Regards



Arne


\

Arne Stolbjerg Drud

ARKI Consulting & Development A/S

Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark

Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: ad…@arki.dk



Fra: gams...@googlegroups.com [mailto:gams...@googlegroups.com] PÃ¥ vegne af Claudia Sonaglio
Sendt: 31. oktober 2011 18:57
Til: gams...@googlegroups.com
Emne: Re: Maximum feasibility tolerance



Arne,



I reduced the monetary values ​​for billions of $ and other variables are rates. The message that now appears is as follows:

** Infeasible solution. Reduced gradient less than tolerance.

you have any suggestions for me to identify the problem?





Regards,

Cláudia







MODEL STATISTICS

BLOCKS OF EQUATIONS 66 SINGLE EQUATIONS 2,016
BLOCKS OF VARIABLES 53 SINGLE VARIABLES 2,011
NON ZERO ELEMENTS 9,791 NON LINEAR N-Z 3,798
DERIVATIVE POOL 6 CONSTANT POOL 197
CODE LENGTH 11,844


GENERATION TIME = 0.343 SECONDS 5 Mb WIN236-236 Mar 30, 2011


EXECUTION TIME = 0.359 SECONDS 5 Mb WIN236-236 Mar 30, 2011

Solution Report SOLVE csonaglio Using NLP From line 1296


S O L V E S U M M A R Y

MODEL csonaglio OBJECTIVE GDPSUM
TYPE NLP DIRECTION MINIMIZE
SOLVER CONOPT FROM LINE 1296

**** SOLVER STATUS 1 Normal Completion
**** MODEL STATUS 5 Locally Infeasible
**** OBJECTIVE VALUE 0.0000

RESOURCE USAGE, LIMIT 136.252 10000.000
ITERATION COUNT, LIMIT 1798 9000
EVALUATION ERRORS 0 0
CONOPTD 0.1 Dec 13, 2010 23.6.4 WIN 23843.23989 VS8 x86/MS Windows


C O N O P T 3 version 3.14W
Copyright (C) ARKI Consulting and Development A/S
Bagsvaerdvej 246 A
DK-2880 Bagsvaerd, Denmark


** Infeasible solution. Reduced gradient less than tolerance.











On Fri, Oct 28, 2011 at 6:44 AM, Arne Stolbjerg Drud wrote:

Claudia:

This is not a problem with the feasibility tolerance. You have very large
variables. You can scale the part of the model with these large variables
(this mean that you measure them in a different unit so the numerical value
becomes smaller). An alternative is to define the option Rtmaxv (mentioned
in the message) to a larger value than the default, for example Rtmaxv =
1.e15. This is done in the conopt.opt option file and you must remember to
tell the solver to use an options file in the usual way.

Good luck

Arne

Arne Stolbjerg Drud
ARKI Consulting & Development A/S
Bagsvaerdvej 246A, DK-2880 Bagsvaerd, Denmark
Phone: (+45) 44 49 03 23, Fax: (+45) 44 49 03 33, email: ad…@arki.dk

-----Oprindelig meddelelse-----
Fra: gams...@googlegroups.com [mailto:gams...@googlegroups.com] PÃ¥ vegne
af Claudia Sonaglio
Sendt: 27. oktober 2011 21:53
Til: gamsworld
Emne: Maximum feasibility tolerance



Hi,
I’m working with a nlp model and the following message appears to
me:

Infeasible solution. A free variable exceeds the allowable range.
Current value is 3.37E+10 and current upper bound (Rtmaxv) is 1.00E
+10

how I could change the Maximum feasibility tolerance?

thank you

–
model solution.PNG