BONMIN#

The COIN Bonmin solver (BONMIN) is an experimental open-source C++ code for solving general MINLP (Mixed Integer NonLinear Programming) problems of the form:

   min     f(x)

s.t.       g_L <= g(x) <= g_U
           x_L <=  x   <= x_U
           x_i in Z for all i in I and,
           x_i in R for all i not in I.

where f(x): R^n --> R, g(x): R^n --> R^m are twice continuously differentiable functions and I is a subset of {1,..,n}.

The algorithms in Bonmin are exact when the functions f and g are convex; in the case where f or g or both are non-convex they are heuristics.

Learn More | Features | Options

Download BONMIN

How to use it#

How to install using amplpy:

# Install Python API for AMPL:
$ python -m pip install amplpy --upgrade

# Install AMPL & solver modules:
$ python -m amplpy.modules install bonmin # install BONMIN

# Activate your license (e.g., free ampl.com/ce or ampl.com/academia licenses):
$ python -m amplpy.modules activate <your-license-uuid>

How to use:

from amplpy import AMPL
ampl = AMPL()
...
ampl.solve(solver="bonmin", bonmin_options="option1=value1 option2=value2")

Learn more about what we have to offer to implement and deploy Optimization in Python.

AMPL APIs give access to AMPL from a programming language:

ampl: option solver bonmin; # change the solver
ampl: option bonmin_options 'option1=value1 option2=value2'; # specify options
ampl: solve; # solve the problem

Resources#

Features#

Algorithms#

Algorithms

bonmin.algorithm: B-BB (NLP-based branch-and-bound, default), B-OA (outer approximation), B-QG (Quesada-Grossmann), B-Hyb (hybrid), B-Ecp, B-iFP. Exact for convex models, heuristic otherwise. Time limit: bonmin.time_limit. Ipopt options (for the NLP subproblems) are also accepted.

ampl: option bonmin_options 'bonmin.algorithm=B-OA bonmin.time_limit=60';

Modeling with MP expressions#

MP expressions: logic, counting, complementarity (MP2NL)

Natively, Bonmin solves smooth nonlinear models with continuous and integer variables. MP’s expressions (==>, or, atmost, count, alldiff, piecewise-linear terms, …) and complements are not part of its native model type: through MP2NL, MP reformulates them into it first.

Call MP2NL as the solver and name Bonmin inside it; MP’s options go into mp2nl_options, Bonmin’s own into bonmin_options.

MP reformulates logic with binary variables, which Bonmin handles as integer variables.

ampl.solve(solver="mp2nl", mp2nl_options="solver=bonmin cvt:compl=1")

ampl: option solver mp2nl;
ampl: option mp2nl_options 'solver=bonmin cvt:compl=1';
ampl: solve;

More: Meta-driver MP2NL.

Solver options#

Full list of solver options:

Retrieving solutions#

The outcome of the last optimization is stored in the AMPL parameter solve_result_num and the relative message in solve_result.

display solve_result_num, solve_result;

BONMIN solve result codes:

Solve result table (AMPL default ranges)
          0- 99 solved: optimal for an optimization problem, feasible for a satisfaction problem
        100-199 solved? solution candidate returned but error likely
        200-299 infeasible
        300-399 unbounded
        400-499 limit
        500-999 failure, no solution returned

Changelog#