def solve(self, *args, **kw):
import networkx as nx
graph = nx.complement(self.graph)
from openopt import STAB
KW = self.__init_kwargs
KW.update(kw)
P = STAB(graph, **KW)
r = P.solve(*args)
return r
def solve(self, *args, **kw):
import networkx as nx
graph = nx.complement(self.graph)
from openopt import STAB
KW = self.__init_kwargs
KW.update(kw)
P = STAB(graph, **KW)
r = P.solve(*args)
return r
'''
Simple OpenOpt graph stability number example;
requires networkx (http://networkx.lanl.gov)
and FuncDesigner installed.
For maximum clique problem you could use STAB on complementary graph, for networkx it's nx.complement(G)
Unlike networkx maximum_independent_set() we search for *exact* solution.
Limitations on time, cputime, "enough" value, basic GUI features are available.
'''
from openopt import STAB
import networkx as nx
G = nx.path_graph(15) # [(0,1), (1,2), (2,3), ..., (13,14)]
# you can use string and other types, e.g.
# G = nx.Graph(); G.add_node('asdf'); G.add_nodes_from(['a1', 'b2','c3']); G.add_edge('qwerty1','qwerty2')
p = STAB(G, includedNodes = [1, 5, 8], excludedNodes = [0, 4, 10])
# or
# p = STAB(G, includedNodes = ['a1', 'c3'], excludedNodes = ['qwerty1'])
# includedNodes and excludedNodes are optional arguments - nodes that must be present or absent in the solution
# solve by BSD- licensed global nonlinear solver interalg (http://openopt.org/interalg):
r = p.solve('interalg', iprint = 0) # set it to -1 to completely suppress interalg output; may not work for glpk, lpSolve
# or solve by MILP solver:
#r = p.solve('lpSolve') # see http://openopt.org/MILP for other MILP solvers available in OpenOpt
# if your solver is cplex or interalg, you can provide some stop criterion,
# e.g. maxTime, maxCPUTime, fEnough etc, for example
# r = p.solve('cplex', maxTime = 100, fEnough = 10)
# i.e. stop if solution with 10 nodes has been obtained or 100 sec were elapsed
# (gplk and lpSolve has no iterfcn connected and thus cannot handle those criterions)
'''
Simple OpenOpt graph stability number example;
requires networkx (http://networkx.lanl.gov)
and FuncDesigner installed.
For maximum clique problem you could use STAB on complementary graph, for networkx it's nx.complement(G)
Unlike networkx maximum_independent_set() we search for *exact* solution.
Limitations on time, cputime, "enough" value, basic GUI features are available.
'''
from openopt import STAB
import networkx as nx
G = nx.path_graph(15) # [(0,1), (1,2), (2,3), ..., (13,14)]
# you can use string and other types, e.g.
# G = nx.Graph(); G.add_node('asdf'); G.add_nodes_from(['a1', 'b2','c3']); G.add_edge('qwerty1','qwerty2')
p = STAB(G, includedNodes=[1, 5, 8], excludedNodes=[0, 4, 10])
# or
# p = STAB(G, includedNodes = ['a1', 'c3'], excludedNodes = ['qwerty1'])
# includedNodes and excludedNodes are optional arguments - nodes that must be present or absent in the solution
# solve by BSD- licensed global nonlinear solver interalg (http://openopt.org/interalg):
r = p.solve(
'interalg', iprint=0
) # set it to -1 to completely suppress interalg output; may not work for glpk, lpSolve
# or solve by MILP solver:
#r = p.solve('lpSolve') # see http://openopt.org/MILP for other MILP solvers available in OpenOpt
# if your solver is cplex or interalg, you can provide some stop criterion,
# e.g. maxTime, maxCPUTime, fEnough etc, for example
# r = p.solve('cplex', maxTime = 100, fEnough = 10)
# i.e. stop if solution with 10 nodes has been obtained or 100 sec were elapsed
