| Yuanyang ZOU |
April 20, 2023 04:00 |
SU2 configuration
I ran my shape optimization Python program directly, and it seems to have started from the task manager, but there is no output in my terminal regarding the iteration information. I want to know if there is a problem with my configuration file or if my shape optimization Python program is not written correctly.Here is my Python program.
Code:
import os, sys, shutil, copy
from optparse import OptionParser
sys.path.append(os.environ['SU2_RUN'])
import SU2
import numpy as np
# -------------------------------------------------------------------
# Main
# -------------------------------------------------------------------
def main():
parser=OptionParser()
parser.add_option("-f", "--file", dest="filename",
help="read config from FILE", metavar="FILE")
parser.add_option("-r", "--name", dest="projectname", default='',
help="try to restart from project file NAME", metavar="NAME")
parser.add_option("-n", "--partitions", dest="partitions", default=1,
help="number of PARTITIONS", metavar="PARTITIONS")
parser.add_option("-g", "--gradient", dest="gradient", default="CONTINUOUS_ADJOINT",
help="Method for computing the GRADIENT (CONTINUOUS_ADJOINT, DISCRETE_ADJOINT, FINDIFF, NONE)", metavar="GRADIENT")
parser.add_option("-o", "--optimization", dest="optimization", default="SLSQP",
help="OPTIMIZATION techique (SLSQP, CG, BFGS, POWELL)", metavar="OPTIMIZATION")
parser.add_option("-q", "--quiet", dest="quiet", default="False",
help="True/False Quiet all SU2 output (optimizer output only)", metavar="QUIET")
parser.add_option("-c", "--n_configs", dest="num_configs", default=1,
help="Number of config files (for weights NN FIML case)", metavar="NUM_CONFIGS")
parser.add_option("-b", "--baseline_nn", dest="weights_file", default=None,
help="Baseline Weights File for Neural Network (If none defaults to random init)", metavar="BASE_NETWORK")
(options, args)=parser.parse_args()
# process inputs
options.partitions = int( options.partitions )
options.quiet = options.quiet.upper() == 'FALSE'
options.gradient = options.gradient.upper()
sys.stdout.write('\n-------------------------------------------------------------------------\n')
sys.stdout.write('| ___ _ _ ___ |\n')
sys.stdout.write('| / __| | | |_ ) Release 5.0.0 \"Raven\" |\n')
sys.stdout.write('| \\__ \\ |_| |/ / |\n')
sys.stdout.write('| |___/\\___//___| Aerodynamic Shape Optimization Script |\n')
sys.stdout.write('| |\n')
sys.stdout.write('-------------------------------------------------------------------------\n')
sys.stdout.write('| SU2 Lead Dev.: Dr. Francisco Palacios, Francisco.D.Palacios@boeing.com|\n')
sys.stdout.write('| Dr. Thomas D. Economon, economon@stanford.edu |\n')
sys.stdout.write('-------------------------------------------------------------------------\n')
sys.stdout.write('| SU2 Developers: |\n')
sys.stdout.write('| - Prof. Juan J. Alonso\'s group at Stanford University. |\n')
sys.stdout.write('| - Prof. Piero Colonna\'s group at Delft University of Technology. |\n')
sys.stdout.write('| - Prof. Nicolas R. Gauger\'s group at Kaiserslautern U. of Technology. |\n')
sys.stdout.write('| - Prof. Alberto Guardone\'s group at Polytechnic University of Milan. |\n')
sys.stdout.write('| - Prof. Rafael Palacios\' group at Imperial College London. |\n')
sys.stdout.write('-------------------------------------------------------------------------\n')
sys.stdout.write('| Copyright (C) 2012-2017 SU2, the open-source CFD code. |\n')
sys.stdout.write('| |\n')
sys.stdout.write('| SU2 is free software; you can redistribute it and/or |\n')
sys.stdout.write('| modify it under the terms of the GNU Lesser General Public |\n')
sys.stdout.write('| License as published by the Free Software Foundation; either |\n')
sys.stdout.write('| version 2.1 of the License, or (at your option) any later version. |\n')
sys.stdout.write('| |\n')
sys.stdout.write('| SU2 is distributed in the hope that it will be useful, |\n')
sys.stdout.write('| but WITHOUT ANY WARRANTY; without even the implied warranty of |\n')
sys.stdout.write('| MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |\n')
sys.stdout.write('| Lesser General Public License for more details. |\n')
sys.stdout.write('| |\n')
sys.stdout.write('| You should have received a copy of the GNU Lesser General Public |\n')
sys.stdout.write('| License along with SU2. If not, see <http://www.gnu.org/licenses/>. |\n')
sys.stdout.write('-------------------------------------------------------------------------\n')
shape_optimization( options.filename ,
options.projectname ,
options.partitions ,
options.gradient ,
options.optimization ,
options.quiet ,
options.num_configs ,
options.weights_file)
#: main()
def shape_optimization( filename ,
projectname = '' ,
partitions = 0 ,
gradient = 'CONTINUOUS_ADJOINT' ,
optimization = 'SLSQP' ,
quiet = False ,
num_configs = 1 ,
weights_file = None):
# Config
config = SU2.io.Config(filename)
#if num_configs > 1 :
# config_vec = config #Use first configuration as default for rest of script JRH 09272018
# for i in range(2,num_configs) :
# config_vec.append(SU2.io.Config(str(i)+'.cfg'))
# config_vec[i].NUMBER_PART = partitions
# config_vec[i].GRADIENT_METHOD = 'gradient'
# if quiet: config_vec[i].CONSOLE = 'CONCISE'
config.NUMBER_PART = partitions
if quiet: config.CONSOLE = 'CONCISE'
config.GRADIENT_METHOD = gradient
its = int ( config.OPT_ITERATIONS )
accu = float ( config.OPT_ACCURACY )
bound_upper = float ( config.OPT_BOUND_UPPER )
bound_lower = float ( config.OPT_BOUND_LOWER )
def_dv = config.DEFINITION_DV
kind_dv = config.DV_KIND
if config.NPOIN == 0:
n_dv = sum(def_dv['SIZE'])
x0 = [0.0]*n_dv
else :
if config.KIND_TRAIN_NN == 'WEIGHTS':
#ANY CHANGES NEED TO BE MIMICKED IN PROJECTION.PY
if weights_file == None :
layers = int(config.N_HIDDEN_LAYERS)+2
num_nodes = [5] #5 if using bias nodes
for iLayer in range(1,layers-1) :
num_nodes.append(int(config.N_NEURONS))
num_nodes.append(1) #2 if using bias nodes
num_inputs = [5] #5 if using bias nodes
#num_inputs.append(int(5))
for iLayer in range(1,layers) :
num_inputs.append(int(config.N_NEURONS))
n_dv = 0
x0 = []
for iLayer in range(1,layers) :
for iNode in range(num_nodes[iLayer]) :
for iInput in range(num_inputs[iLayer-1]) :
#x0.append(np.random.randn()/np.sqrt(5.0)/10.0+0.1)
#x0.append(0.0)
if iLayer > 0 : #JRH 09232018 - Removing input layer weights from costly computation
x0.append(np.random.randn()*np.sqrt(1.0/num_inputs[iLayer-1]))
n_dv+=1
sys.stdout.write('Total Number of Weights Set to ' + str(n_dv) + ' in shape_optimization.py\n')
else :
sys.stdout.write('Reading weights from' + str(weights_file) + '\n')
weights = open(weights_file)
x0 = []
lines = weights.readlines() #i am reading lines here
n_dv = 0
for line in lines: #taking each line
x0.append(float(line)) #converting string to float
n_dv = n_dv+1
#sys.stdout.write('n_dv ' + str(x0[n_dv-1]) +'\n')
sys.stdout.write('Total Number of Weights Set to ' + str(n_dv) + ' in shape_optimization.py \n')
else:
if weights_file == None :
n_dv = int (config.NPOIN)
x0 = [float(config.DV_VALUE[0])-1.0]*n_dv # initial design
else :
sys.stdout.write('Reading betas from' + str(weights_file) + '\n')
weights = open(weights_file)
x0 = []
lines = weights.readlines() #i am reading lines here
n_dv = 0
for line in lines: #taking each line
x0.append(float(line)) #converting string to float
n_dv = n_dv+1
#sys.stdout.write('n_dv ' + str(x0[n_dv-1]) +'\n')
sys.stdout.write('Total Number of Betas Set to ' + str(n_dv) + ' in shape_optimization.py \n')
xb_low = [float(bound_lower)]*n_dv # lower dv bound
xb_up = [float(bound_upper)]*n_dv # upper dv bound
xb = zip(xb_low,xb_up) # design bounds
sys.stdout.write('NPOIN = %s\n' % config.NPOIN)
# State
state = SU2.io.State()
state.find_files(config)
# Project
if os.path.exists(projectname):
project = SU2.io.load_data(projectname)
project.config = config
else:
#if num_configs > 1 :
#project = SU2.opt.Project(config,state, config_vector = config_vec)
#else :
project = SU2.opt.Project(config,state)
# Optimize
if optimization == 'SLSQP':
SU2.opt.SLSQP(project,x0,xb,its,accu)
if optimization == 'CG':
SU2.opt.CG(project,x0,xb,its,accu)
if optimization == 'BFGS':
SU2.opt.BFGS(project,x0,xb,its,accu)
if optimization == 'BFGSJ':
SU2.opt.BFGSJ(project,x0,xb,its,accu)
if optimization == 'STEEPJ':
SU2.opt.STEEPJ(project,x0,xb,its,accu)
if optimization == 'POWELL':
SU2.opt.POWELL(project,x0,xb,its,accu)
if optimization == 'BFGSG':
SU2.opt.BFGSG(project,x0,xb,its,accu)
if optimization == 'STEEP':
SU2.opt.STEEP(project,x0,xb,its,accu)
if optimization == 'TNC':
SU2.opt.TNC(project,x0,xb,its,accu)
# rename project file
if projectname:
shutil.move('project.pkl',projectname)
return project
#: shape_optimization()
# -------------------------------------------------------------------
# Run Main Program
# -------------------------------------------------------------------
# this is only accessed if running from command prompt
if __name__ == '__main__':
main()
|
