import json
from algos import OptAlg
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
if not os.path.exists('./results'):
os.makedirs('./results')
print("Train LPMC_DC_L with LS-ABS and hybrid-inv")
model = LPMC_DC(data_folder, file='12_13_14.csv')
ioa = OptAlg(alg_type='LS-ABS', direction='hybrid-inv')
res = {'time': [], 'LL': [], 'epochs': []}
for i in range(20):
tmp = model.optimize(
ioa, **{
'verbose': False,
'max_epochs': 1000,
'batch': 1000
})
res['time'].append(tmp['opti_time'])
import json
from algos import OptAlg
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
if not os.path.exists('./results'):
os.makedirs('./results')
print("Train LPMC_DC_M with LS-ABS and grad")
model = LPMC_DC(data_folder)
ioa = OptAlg(alg_type='LS-ABS', direction='grad')
res = {'time': [], 'LL': [], 'epochs': []}
for i in range(20):
tmp = model.optimize(
ioa, **{
'verbose': False,
'max_epochs': 1000,
'batch': 1000
})
res['time'].append(tmp['opti_time'])
from algos import OptAlg
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
parameters = {''}
if not os.path.exists('./results'):
os.makedirs('./results')
print("Testing parameters for LPMC_RR_L")
model = LPMC_DC(data_folder, file='12_13_14.csv')
ioa = OptAlg(alg_type='LS-ABS', direction='hybrid-inv')
base_param = {'perc_hybrid': 30, 'thresh_upd': 1, 'count_upd': 2, 'window': 10, 'factor_upd': 2}
main_params = {'verbose': False, 'max_epochs': 1000, 'batch': 1000}
main_params.update(base_param)
draws = 10
res = {}
print("Start with perc_hybrid")
param_ph = [0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100]
import os
import time
import json
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
if not os.path.exists('./results'):
os.makedirs('./results')
print("Train LPMC_DC with current Biogeme optimization algorithm.")
model = LPMC_DC(data_folder, file='12.csv')
res = {'time': [], 'LL': [], 'epochs': []}
for i in range(20):
results = model.biogeme.estimate()
res['time'].append(model.biogeme.optimizationTime.total_seconds())
res['LL'].append(results.data.logLike)
res['epochs'].append(model.biogeme.numberOfIterations)
with open('results/biogeme_bfgs.json', 'w') as outfile:
json.dump(res, outfile)
# Delete pickle files
import os
import time
import json
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
if not os.path.exists('./results'):
os.makedirs('./results')
print("Train LPMC_DC with current Biogeme optimization algorithm.")
model = LPMC_DC(data_folder, file='12_13_14.csv')
res = {'time': [], 'LL': [], 'epochs': []}
for i in range(20):
results = model.biogeme.estimate()
res['time'].append(model.biogeme.optimizationTime.total_seconds())
res['LL'].append(results.data.logLike)
res['epochs'].append(model.biogeme.numberOfIterations)
with open('results/biogeme_bfgs.json', 'w') as outfile:
json.dump(res, outfile)
# Delete pickle files
import os
import time
import json
from models import LPMC_DC
data_folder = '../../../data/'
if __name__ == "__main__":
if not os.path.exists('./results'):
os.makedirs('./results')
print("Train LPMC_DC with current Biogeme optimization algorithm.")
model = LPMC_DC(data_folder)
res = {'time': [], 'LL': [], 'epochs': []}
for i in range(20):
results = model.biogeme.estimate()
res['time'].append(model.biogeme.optimizationTime.total_seconds())
res['LL'].append(results.data.logLike)
res['epochs'].append(model.biogeme.numberOfIterations)
with open('results/biogeme_bfgs.json', 'w') as outfile:
json.dump(res, outfile)
# Delete pickle files
