Error "Bailing out after 10 warnings"

[Jennifer Arellana Guzman]

Hello all,
I am making a model in which I have to get the result of the objective function the number of times (2^n)-1, which would represent the objective function for each possible combination of the n individuals. I ran the model with n=5 and it should have given me 31 results, but when I was going for result 16 it was interrupted and I got the following:

Bailing out after 10 warnings.
presolve: constraint Oferta['O5','P5','T2'] cannot hold:
    body >= 295 cannot be <= 1.38607e-09; difference = 295

Does anyone know why I got this error?

I appreciate any help and thank you for your time.

[AMPL Google Group]

The "presolve" message is telling you that AMPL has not sent your problem to the solver, because AMPL's presolve phase has analyzed the constraints (including the bounds on the variables) and has determined that no feasible solution is possible. However, I see before that the message "Bailing out after 10 warnings." So to get help, you should run the model again, and reply with a copy of all the messages from AMPL (including those 10 warnings). Also if you can attach your files, that might help in explaining what is going on.

--
Robert Fourer
am...@googlegroups.com

{#HS:1532152442-104735#}

[Jennifer Arellana Guzman]

Dear Mr. Robert,

Unfortunately I can't attach the documents but here I show all the warnings below:

- First I get these warnings before I get the first results:

presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
    all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
    no variables, but lower bound = 1, upper = Infinity

- When it returns result number 15, these warnings appear:

presolve, constraint V_binaria2['A1','C2','P4','R1','T1']:
    all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
    all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
    no variables, but lower bound = 1, upper = Infinity
presolve, constraint V_binaria2['A1','C2','P4','R1','T1']:

    no variables, but lower bound = 1, upper = Infinity

- Finally, when I get the result number 16 I get this last warning and the process is interrupted:

Bailing out after 10 warnings.
presolve: constraint Oferta['O5','P5','T2'] cannot hold:
    body >= 295 cannot be <= 1.38607e-09; difference = 295

I appreciate any help and thank you for your time.

[AMPL Google Group]

I suggest that you start by concentrating on the warnings for the first problem. Stop your run after the "solve" for the first problem, so that you can use some AMPL commands to see what is happening.

The warnings are telling you that the first problem has no feasible solution: there is no way to assign values to the variables that satisfy all of the constraints. Also the messages point to the constraint V_binaria2['A1','C2','P3','R1','T2'] as the cause of this difficulty; so use the following command to look at the expression that AMPL generated for this constraint:

expand V_binaria2['A1','C2','P3','R1','T2'];

Then display the lower and upper bounds deduced by presolve for each of the variables in this constraint, using commands like

display <varname>.lb, <varname>.ub;

where <varname> is replaced by the name of each variable as shown in the "expand" output. You will find that the lower bound equals the upper bound for these variables, meaning that they have only one possible value, but when you put those values into V_binaria2['A1','C2','P3','R1','T2'], that constraint is not satisfied.

Finally, you will need to study your model and data to determine why each variable in V_binaria2['A1','C2','P3','R1','T2'] has only one possible value, and then you will have to figure out how to fix this situation. Maybe a constraint is not written correctly, or maybe the data does not allow any feasible solutions. Since I have not seen the model and data, however, I cannot offer more specific advice than this.

--
Robert Fourer
am...@googlegroups.com

{#HS:1532152442-104735#}

On Thu, Jun 3, 2021 at 10:01 PM UTC, AMPL Modeling Language <am...@googlegroups.com> wrote:

Dear Mr. Robert,
Unfortunately I can't attach the documents but here I show all the warnings below:

- First I get these warnings before I get the first results:

presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
no variables, but lower bound = 1, upper = Infinity

- When it returns result number 15, these warnings appear:

presolve, constraint V_binaria2['A1','C2','P4','R1','T1']:
all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
all variables eliminated, but lower bound = 1 > 0
presolve, constraint V_binaria2['A1','C2','P3','R1','T2']:
no variables, but lower bound = 1, upper = Infinity
presolve, constraint V_binaria2['A1','C2','P4','R1','T1']:
no variables, but lower bound = 1, upper = Infinity

- Finally, when I get the result number 16 I get this last warning and the process is interrupted:

Bailing out after 10 warnings.
presolve: constraint Oferta['O5','P5','T2'] cannot hold:
body >= 295 cannot be <= 1.38607e-09; difference = 295

I appreciate any help and thank you for your time.

On Thu, Jun 3, 2021 at 3:59 PM UTC, AMPL Google Group <am...@googlegroups.com> wrote:

The "presolve" message is telling you that AMPL has not sent your problem to the solver, because AMPL's presolve phase has analyzed the constraints (including the bounds on the variables) and has determined that no feasible solution is possible. However, I see before that the message "Bailing out after 10 warnings." So to get help, you should run the model again, and reply with a copy of all the messages from AMPL (including those 10 warnings). Also if you can attach your files, that might help in explaining what is going on.

--
Robert Fourer
am...@googlegroups.com