bench 1 - interalg

from FuncDesigner import *
from openopt import *

a, b, c, d, e, f = oovars('a', 'b', 'c', 'd', 'e', 'f')
F =  a/(b+1e-15) + c/(d+1e-15) + e/(f+1e-15) + f/(a+1e-15)
startPoint = {a:0.51, b:0.52, c:0.53, d:0.54, e:0.55, f:0.56}
constraints = (a>0, a<1, b>0, b<1, c>0, c<1, d>0, d<1, e>0, e<1, f>0, f<1)
p = GLP(F, startPoint, fTol = 1e-9, iterObjFunTextFormat='%0.9e' , iprint = 10, constraints = constraints)
r = p.solve('interalg')

Output:
--------------------------------------------------
solver: interalg_0.17   problem: unnamed    type: GLP
 iter    objFunVal
    0  4.042432785e+00
   10  1.666666667e+00
   20  2.916666667e-01
   30  3.059895833e-02
   40  2.685546860e-03
   50  6.515979179e-03
   60  3.580725299e-03
   70  1.009040436e-03
   80  1.168485433e-04
   90  1.206315536e-05
  100  1.190005497e-06
  110  1.487507163e-07
  120  1.512797509e-08
  130  1.113102992e-09
OpenOpt info: Solution with required tolerance 1.0e-09
 is guarantied (obtained precision: 4.470e-10)
  133  4.470494630e-10
istop: 1000 (optimal solution obtained)
Solver:   Time Elapsed = 28.67  CPU Time Elapsed = 27.92
objFunValue: 4.4704946e-10 (feasible, MaxResidual = 0)