2D permeability field calibration

import numpy as np

from HydrOpTop.Functions import Reference_Liquid_Head #objective
from HydrOpTop.Functions import Volume_Percentage #constrain
from HydrOpTop.Materials import Log_SIMP
from HydrOpTop.Crafter import Steady_State_Crafter
from HydrOpTop.Filters import Density_Filter
from HydrOpTop.Solvers import PFLOTRAN

#create PFLOTRAN simulation object
pflotranin = "pflotran.in"
sim = PFLOTRAN(pflotranin)

#get cell ids in the region to optimize and parametrize permeability
#same name than in pflotran input file
perm = Log_SIMP(cell_ids_to_parametrize="all", property_name="PERMEABILITY", bounds=[1e-14, 1e-12], power=1)

#define cost function as sum of the head in the pit
cell_ids = [444, 789, 920, 1030, 1339]
head = [230, 250, 227, 146, 210]
cf = Reference_Liquid_Head(head, cell_ids, norm=2)

#define maximum volume constrains
#max_vol = Volume_Percentage("parametrized_cell")
#max_vol.constraint_tol = 0.2

filter_ = Density_Filter(10.)

#craft optimization problem
#i.e. create function to optimize, initiate IO array in classes...
crafted_problem = Steady_State_Crafter(cf, sim, [perm], [], [filter_])
crafted_problem.IO.output_every_iteration(1)
crafted_problem.IO.output_gradient()
crafted_problem.IO.output_material_properties()
crafted_problem.IO.define_output_format("vtu")

#initialize optimizer
p = np.zeros(crafted_problem.get_problem_size()) + 0.5
#p = 0.999*np.random.rand(crafted_problem.get_problem_size())+0.001
p_opt = crafted_problem.optimize(optimizer="nlopt-mma", action="minimize", max_it=100, ftol=0.0001, initial_guess=p)