Python cplex.Cplex方法代碼示例

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def __init__(self, cplex=None):
        if not _HAS_CPLEX:
            raise NameError('CPLEX is not installed. '
                            'See https://www.ibm.com/support/knowledgecenter/'
                            'SSSA5P_12.8.0/ilog.odms.studio.help/Optimization_Studio/'
                            'topics/COS_home.html')

        if cplex:
            self._model = Cplex(cplex._model)
        else:
            self._model = Cplex()

        self._init_lin()
        # to avoid a variable with index 0
        self._model.variables.add(names=['_dummy_'], types=[self._model.variables.type.continuous])
        self._var_id = {'_dummy_': 0} 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def solve_with_cplex(path):
    import cplex
    problem = cplex.Cplex()
    problem.parameters.read.datacheck.set(0)
    problem.set_results_stream("%s_cplex.log" % path)
    problem.read(("%s.lp" % path).encode("utf8")) # unicode conversion required by CPLEX (undocumented)
    problem.solve()
    if problem.solution.get_status() != 103: # No solution exists
        return {
            "distances": problem.solution.get_objective_value(),
            "names": problem.variables.get_names(),
            "assigned": problem.solution.get_values(),
        } 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def add_mip_starts(mip, indices, pool, max_mip_starts = float('inf'), mip_start_effort_level = 4):
    """

    Parameters
    ----------
    mip - RiskSLIM surrogate MIP
    indices - indices of RiskSLIM surrogate MIP
    pool - solution pool
    max_mip_starts - max number of mip starts to add (optional; default is add all)
    mip_start_effort_level - effort that CPLEX will spend trying to fix (optional; default is 4)

    Returns
    -------

    """
    # todo remove suboptimal using pool filter
    assert isinstance(mip, Cplex)

    try:
        obj_cutoff = mip.parameters.mip.tolerances.uppercutoff.get()
    except:
        obj_cutoff = float('inf')

    pool = pool.distinct().sort()

    n_added = 0
    for objval, rho in zip(pool.objvals, pool.solutions):
        if np.less_equal(objval, obj_cutoff):
            mip_start_name = "mip_start_" + str(n_added)
            mip_start_obj, _ = convert_to_risk_slim_cplex_solution(rho = rho, indices = indices, objval = objval)
            mip_start_obj = cast_mip_start(mip_start_obj, mip)
            mip.MIP_starts.add(mip_start_obj, mip_start_effort_level, mip_start_name)
            n_added += 1

        if n_added >= max_mip_starts:
            break

    return mip 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def cast_mip_start(mip_start, cpx):
    """
    casts the solution values and indices in a Cplex SparsePair

    Parameters
    ----------
    mip_start cplex SparsePair
    cpx Cplex

    Returns
    -------
    Cplex SparsePair where the indices are integers and the values for each variable match the variable type specified in CPLEX Object
    """

    assert isinstance(cpx, Cplex)
    assert isinstance(mip_start, SparsePair)
    vals = list(mip_start.val)
    idx = np.array(list(mip_start.ind), dtype = int).tolist()
    types = cpx.variables.get_types(idx)

    for j, t in enumerate(types):
        if t in ['B', 'I']:
            vals[j] = int(vals[j])
        elif t in ['C']:
            vals[j] = float(vals[j])

    return SparsePair(ind = idx, val = vals) 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def findSolutionValues(self, lp):
            lp.cplex_status = lp.solverModel.solution.get_status()
            lp.status = self.CplexLpStatus.get(lp.cplex_status, LpStatusUndefined)
            var_names = [var.name for var in lp.variables()]
            con_names = [con for con in lp.constraints]
            try:
                objectiveValue = lp.solverModel.solution.get_objective_value()
                variablevalues = dict(zip(var_names, lp.solverModel.solution.get_values(var_names)))
                lp.assignVarsVals(variablevalues)
                constraintslackvalues = dict(zip(con_names, lp.solverModel.solution.get_linear_slacks(con_names)))
                lp.assignConsSlack(constraintslackvalues)
                if lp.solverModel.get_problem_type == cplex.Cplex.problem_type.LP:
                    variabledjvalues = dict(zip(var_names, lp.solverModel.solution.get_reduced_costs(var_names)))
                    lp.assignVarsDj(variabledjvalues)
                    constraintpivalues = dict(zip(con_names, lp.solverModel.solution.get_dual_values(con_names)))
                    lp.assignConsPi(constraintpivalues)
            except cplex.exceptions.CplexSolverError:
                #raises this error when there is no solution
                pass
            #put pi and slack variables against the constraints
            #TODO: clear up the name of self.n2c
            if self.msg:
                print("Cplex status=", lp.cplex_status)
            lp.resolveOK = True
            for var in lp.variables():
                var.isModified = False
            return lp.status 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def cplex_solution(self):
        """ cplex solution """

        # refactoring
        rho = self.rho
        n = self.n
        q = self.q

        my_obj = list(rho.reshape(1, n ** 2)[0]) + [0. for x in range(0, n)]
        my_ub = [1 for x in range(0, n ** 2 + n)]
        my_lb = [0 for x in range(0, n ** 2 + n)]
        my_ctype = "".join(['I' for x in range(0, n ** 2 + n)])

        my_rhs = [q] + [1 for x in range(0, n)] + \
                 [0 for x in range(0, n)] + [0.1 for x in range(0, n ** 2)]
        my_sense = "".join(['E' for x in range(0, 1 + n)]) + \
                   "".join(['E' for x in range(0, n)]) + \
                   "".join(['L' for x in range(0, n ** 2)])

        try:
            my_prob = cplex.Cplex()
            self._populate_by_row(my_prob, my_obj, my_ub, my_lb, my_ctype, my_sense, my_rhs)

            my_prob.solve()

        except Exception as ex:  # pylint: disable=broad-except
            print(str(ex))
            return None, None

        x = my_prob.solution.get_values()
        x = np.array(x)
        cost = my_prob.solution.get_objective_value()

        return x, cost 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def createCplex( **params):
        """Create and return a :class:`cplex.Cplex` instance with disabled debugging output. Keyword
        args are used to set parameters.

        CPLEX parameters can be set by using their name in the python interface, excluding the
        leading ``cplex.parameters.``, as key (e.g. ``workmem``, ``mip.strategy``).
        """
        import cplex
        cpx = cplex.Cplex()
        stream = None
        if params.get('debug', False):
            stream = sys.stderr
        if 'debug' in params:
            del params['debug']
        cpx.set_results_stream(stream)
        cpx.set_warning_stream(stream)
        cpx.set_error_stream(stream)
        if 'version' in params:
            assert cpx.get_version() == params.pop('version')

        for arg, val in params.items():
            parts = arg.split('.')
            param = cpx.parameters
            for part in parts:
                param = getattr(param, part)
            param.set(val)
        return cpx 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def checkKeyboardInterrupt(cpx):
    """Checks the solution status of the given :class:`cplex.Cplex` instance for keyboard
    interrupts, and raises a :class:`KeyboardInterrupt` exception in that case.
    """
    import cplex
    if cpx.solution.get_status() in (cplex._internal._constants.CPX_STAT_ABORT_USER,
                                     cplex._internal._constants.CPXMIP_ABORT_FEAS,
                                     cplex._internal._constants.CPXMIP_ABORT_INFEAS):
        raise KeyboardInterrupt() 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def copy_cplex(cpx):
    cpx_copy = Cplex(cpx)
    cpx_parameters = cpx.parameters.get_changed()
    for (pname, pvalue) in cpx_parameters:
        phandle = reduce(getattr, str(pname).split("."), cpx_copy)
        phandle.set(pvalue)
    return cpx_copy


# Building 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def run_and_read_cplex(n, problem_fn, solution_fn, solver_logfile,
                        solver_options, warmstart=None, store_basis=True):
    """
    Solving function. Reads the linear problem file and passes it to the cplex
    solver. If the solution is sucessful it returns variable solutions and
    constraint dual values. Cplex must be installed for using this function

    """
    if find_spec('cplex') is None:
        raise ModuleNotFoundError("Optional dependency 'cplex' not found."
           "Install via 'conda install -c ibmdecisionoptimization cplex' "
           "or 'pip install cplex'")
    import cplex
    m = cplex.Cplex()
    out = m.set_log_stream(solver_logfile)
    if solver_options is not None:
        for key, value in solver_options.items():
            param = m.parameters
            for key_layer in key.split("."):
                param = getattr(param, key_layer)
            param.set(value)
    m.read(problem_fn)
    if warmstart:
        m.start.read_basis(warmstart)
    m.solve()
    is_lp = m.problem_type[m.get_problem_type()] == 'LP'

    termination_condition = m.solution.get_status_string()
    if 'optimal' in termination_condition:
        status = 'ok'
        termination_condition = 'optimal'
    else:
        status = 'warning'

    if (status == 'ok') and store_basis and is_lp:
        n.basis_fn = solution_fn.replace('.sol', '.bas')
        try:
            m.solution.basis.write(n.basis_fn)
        except cplex.exceptions.errors.CplexSolverError:
            logger.info('No model basis stored')
            del n.basis_fn

    objective = m.solution.get_objective_value()
    variables_sol = pd.Series(m.solution.get_values(), m.variables.get_names())\
                      .pipe(set_int_index)
    if is_lp:
        constraints_dual = pd.Series(m.solution.get_dual_values(),
                         m.linear_constraints.get_names()).pipe(set_int_index)
    else:
        logger.warning("Shadow prices of MILP couldn't be parsed")
        constraints_dual = pd.Series(index=m.linear_constraints.get_names())\
                             .pipe(set_int_index)
    del m
    return (status, termination_condition, variables_sol, constraints_dual,
            objective) 

# 需要導入模塊: import cplex [as 別名]
# 或者: from cplex import Cplex [as 別名]
def findSolutionValues(self, lp):
            CplexLpStatus = {lp.solverModel.solution.status.MIP_optimal: LpStatusOptimal,
                             lp.solverModel.solution.status.optimal: LpStatusOptimal,
                             lp.solverModel.solution.status.optimal_tolerance: LpStatusOptimal,
                             lp.solverModel.solution.status.infeasible: LpStatusInfeasible,
                             lp.solverModel.solution.status.infeasible_or_unbounded:  LpStatusInfeasible,
                             lp.solverModel.solution.status.MIP_infeasible: LpStatusInfeasible,
                             lp.solverModel.solution.status.MIP_infeasible_or_unbounded:  LpStatusInfeasible,
                             lp.solverModel.solution.status.unbounded: LpStatusUnbounded,
                             lp.solverModel.solution.status.MIP_unbounded: LpStatusUnbounded,
                             lp.solverModel.solution.status.abort_dual_obj_limit: LpStatusNotSolved,
                             lp.solverModel.solution.status.abort_iteration_limit: LpStatusNotSolved,
                             lp.solverModel.solution.status.abort_obj_limit: LpStatusNotSolved,
                             lp.solverModel.solution.status.abort_relaxed: LpStatusNotSolved,
                             lp.solverModel.solution.status.abort_time_limit: LpStatusNotSolved,
                             lp.solverModel.solution.status.abort_user: LpStatusNotSolved}
            lp.cplex_status = lp.solverModel.solution.get_status()
            lp.status = CplexLpStatus.get(lp.cplex_status, LpStatusUndefined)
            var_names = [var.name for var in lp.variables()]
            con_names = [con for con in lp.constraints]
            try:
                objectiveValue = lp.solverModel.solution.get_objective_value()
                variablevalues = dict(zip(var_names, lp.solverModel.solution.get_values(var_names)))
                lp.assignVarsVals(variablevalues)
                constraintslackvalues = dict(zip(con_names, lp.solverModel.solution.get_linear_slacks(con_names)))
                lp.assignConsSlack(constraintslackvalues)
                if lp.solverModel.get_problem_type == cplex.Cplex.problem_type.LP:
                    variabledjvalues = dict(zip(var_names, lp.solverModel.solution.get_reduced_costs(var_names)))
                    lp.assignVarsDj(variabledjvalues)
                    constraintpivalues = dict(zip(con_names, lp.solverModel.solution.get_dual_values(con_names)))
                    lp.assignConsPi(constraintpivalues)
            except cplex.exceptions.CplexSolverError:
                #raises this error when there is no solution
                pass
            #put pi and slack variables against the constraints
            #TODO: clear up the name of self.n2c
            if self.msg:
                print("Cplex status=", lp.cplex_status)
            lp.resolveOK = True
            for var in lp.variables():
                var.isModified = False
            return lp.status