Assisted specification¶

Example of the estimation of several versions of the model using assisted specification algorithm. The catalog of specifications is defined in Specification of a catalog of models . Compared to 21a. Assisted specification, the number of specifications exceeds the maximum limit, so a heuristic is applied. See Bierlaire and Ortelli, 2023 for a detailed description of the use of the assisted specification algorithm.

Michel Bierlaire, EPFL Sat Jun 28 2025, 12:25:12

import biogeme.biogeme_logging as blog
from biogeme.assisted import AssistedSpecification
from biogeme.catalog import count_number_of_specifications
from biogeme.multiobjectives import aic_bic_dimension
from biogeme.results_processing import compile_estimation_results

from plot_b22b_multiple_models_spec import PARETO_FILE_NAME, the_biogeme

logger = blog.get_screen_logger(blog.INFO)
logger.info('Example b22multiple_models')
Example b22multiple_models
nbr = count_number_of_specifications(the_biogeme.log_like)
if nbr is None:
    print('There are too many possible specifications to be enumerated')
else:
    print(f'There are {nbr} possible specifications')
There are 504 possible specifications

Creation of the object capturing the assisted specification algorithm. Its constructor takes three arguments:

  • the biogeme object containing the specifications and the database,

  • an object defining the objectives to minimize. Here, we use three objectives: AIC, BIC and number of parameters.

  • the name of the file where the estimated are saved, and organized into a Pareto set.

assisted_specification = AssistedSpecification(
    biogeme_object=the_biogeme,
    multi_objectives=aic_bic_dimension,
    pareto_file_name=PARETO_FILE_NAME,
)
Unable to read file b22_multiple_models.pareto. Pareto set empty.

The algorithm is run.

non_dominated_models = assisted_specification.run()
Biogeme parameters read from biogeme.toml.
Model with 4 unknown parameters [max: 50]
*** Estimate b07everything_000185
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.92           -0.67           -0.88           -0.49      5.4e+03      0.041       10      1.1   ++
    1           -0.73            -1.2              -1           -0.18      5.3e+03     0.0072    1e+02      1.1   ++
    2            -0.7            -1.3            -1.1           -0.16      5.3e+03    0.00018    1e+03        1   ++
    3            -0.7            -1.3            -1.1           -0.16      5.3e+03    1.1e-07    1e+03        1   ++
default_specification=asc:no_seg;train_cost_catalog:linear;train_headway_catalog:without_headway;train_tt_catalog:linear
The number of possible specifications [504] exceeds the maximum number [100]. A heuristic algorithm is applied.
*** VNS ***
asc:no_seg;train_cost_catalog:linear;train_headway_catalog:without_headway;train_tt_catalog:linear [10670.504013832326, np.float64(10697.78385743747), 4]
Initial pareto: 1
Attempt 0/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000186
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC         asc_car beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0              -1              -1               0               0               0               0               0          -0.079               0               0      5.7e+03      0.036       10      1.1   ++
    1           -0.53            -2.7               0               0               0               0               0            0.14               0               0      5.6e+03      0.012    1e+02      1.1   ++
    2           -0.42              -3               0               0               0               0               0            0.19               0               0      5.6e+03    0.00049    1e+03        1   ++
    3           -0.42              -3               0               0               0               0               0            0.19               0               0      5.6e+03    6.5e-07    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000187
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.88            0.29         -0.0088           -0.77              -1           -0.19           -0.18          -0.056      5.4e+03      0.038       10      1.1   ++
    1            -1.4             0.9            0.75            -1.1            -2.2           -0.24            0.19          -0.068      5.2e+03      0.009    1e+02      1.1   ++
    2            -1.6             1.1            0.95            -1.2            -2.4           -0.24             0.2           -0.19      5.2e+03    0.00085    1e+03        1   ++
    3            -1.6             1.2            0.98            -1.2            -2.4           -0.24             0.2            -0.2      5.2e+03    9.5e-06    1e+04        1   ++
    4            -1.6             1.2            0.98            -1.2            -2.4           -0.24             0.2            -0.2      5.2e+03    1.2e-09    1e+04        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:LUGGAGE;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:linear [10415.149498177643, np.float64(10469.709185387932), 8]
Attempt 1/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000188
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.93             0.5              -1               0               0               0         -0.0031               0               0           -0.65           -0.25               0               0      5.6e+03        2.7       10        1   ++
    1            -0.6             1.8            -2.8               0               0               0          -0.005               0               0            0.11            -1.1               0               0      5.2e+03       0.68    1e+02        1   ++
    2           -0.63             1.8              -3               0               0               0         -0.0059               0               0           0.094            -1.4               0               0      5.2e+03      0.036    1e+03        1   ++
    3           -0.63             1.8              -3               0               0               0          -0.006               0               0           0.094            -1.4               0               0      5.2e+03    0.00017    1e+04        1   ++
    4           -0.63             1.8              -3               0               0               0          -0.006               0               0           0.094            -1.4               0               0      5.2e+03    3.4e-09    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000189
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost     lambda_cost       b_headway         asc_car     Function    Relgrad   Radius      Rho
    0           -0.93              -1           -0.55               1         -0.0032           -0.77      5.6e+03        2.4       10        1   ++
    1           -0.35            -2.9            -1.1            0.23         -0.0053           -0.23      5.4e+03       0.17    1e+02     0.99   ++
    2           -0.33            -3.1            -1.1            0.64         -0.0053           -0.23      5.3e+03      0.005    1e+02     0.85    +
    3           -0.32            -3.1            -1.1            0.59         -0.0054           -0.22      5.3e+03    0.00053    1e+03        1   ++
    4           -0.32            -3.1            -1.1            0.59         -0.0054           -0.22      5.3e+03    1.7e-05    1e+04        1   ++
    5           -0.32            -3.1            -1.1            0.59         -0.0054           -0.22      5.3e+03    1.1e-06    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000190
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost     lambda_cost beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08       10        1   ++
    1              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08        5 -8.3e+05    -
    2              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08      2.5      -42    -
    3            -1.5               0               0               0            -1.7            -0.5               0               0            -1.3               0               0      5.9e+03      0.082      2.5      0.2    +
    4            -1.5               0               0               0            -1.7            -0.5               0               0            -1.3               0               0      5.9e+03      0.082        1    -0.66    -
    5            -1.4               0               0               0           -0.71           -0.45               0               0           -0.58               0               0      5.7e+03      0.026        1     0.82    +
    6            -1.7               0               0               0            -1.4            0.59               0               0           -0.93               0               0      5.7e+03      0.026        1     0.38    +
    7            -1.7               0               0               0            -1.1            0.38               0               0           -0.88               0               0      5.7e+03     0.0015       10      1.1   ++
    8            -1.7               0               0               0            -1.1            0.17               0               0           -0.88               0               0      5.7e+03     0.0012    1e+02     0.98   ++
    9            -1.7               0               0               0            -1.1            0.17               0               0           -0.88               0               0      5.7e+03    3.3e-06    1e+02        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000191
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03          3       10      1.1   ++
    1      5.3e+03       0.43    1e+02      1.1   ++
    2      5.3e+03      0.038    1e+03        1   ++
    3      5.3e+03    0.00034    1e+04        1   ++
    4      5.3e+03    3.1e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Considering neighbor 4/20 for current solution
Attempt 2/100
Considering neighbor 0/20 for current solution
Attempt 3/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000192
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.5e+03      0.039       10        1   ++
    1      5.4e+03     0.0087    1e+02      1.1   ++
    2      5.4e+03    0.00053    1e+03      1.1   ++
    3      5.4e+03    4.5e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000193
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.54           -0.84            0.24         -0.0055              -1           -0.91               1         -0.0017           -0.49           -0.18           -0.22          -0.058      5.4e+03        2.6       10      1.1   ++
    1           -0.22            -1.1            0.75            0.34            -2.7            -1.3            0.24         -0.0049           -0.56            0.25           -0.02           -0.39      5.1e+03       0.48    1e+02        1   ++
    2           -0.23            -1.1            0.93            0.49              -3            -1.2            0.46         -0.0058           -0.51            0.29          -0.051           -0.49      5.1e+03       0.05    1e+03      1.1   ++
    3           -0.24            -1.1            0.96            0.51              -3            -1.1            0.56         -0.0059           -0.51            0.29          -0.051            -0.5      5.1e+03    0.00073    1e+04        1   ++
    4           -0.24            -1.1            0.96            0.51              -3            -1.1            0.56         -0.0059           -0.51            0.29          -0.051            -0.5      5.1e+03    3.6e-06    1e+04        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:MALE-LUGGAGE;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:sqrt [10279.062298118724, np.float64(10360.901828934157), 12]
Attempt 4/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000194
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.86            0.79           -0.82              -1               1           -0.48           -0.37      5.3e+03       0.04       10      1.1   ++
    1            -1.4               2              -1            -1.9           -0.36           -0.45            -1.5      5.1e+03      0.054       10     0.67    +
    2            -1.5             2.1           -0.98            -1.3            -0.3           -0.34            -1.8        5e+03     0.0023    1e+02        1   ++
    3            -1.5             2.1              -1            -1.6            0.14           -0.38            -1.8        5e+03     0.0071    1e+02     0.82    +
    4            -1.5             2.1            -1.1            -1.5            0.12           -0.35            -1.8        5e+03    0.00014    1e+03     0.99   ++
    5            -1.5             2.1            -1.1            -1.5            0.12           -0.35            -1.8        5e+03    1.3e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:boxcox;train_headway_catalog:without_headway;train_tt_catalog:linear [10019.113148838624, np.float64(10066.852875147626), 7]
Attempt 5/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000195
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03      0.038       10      1.1   ++
    1      5.3e+03      0.013    1e+02      1.1   ++
    2      5.3e+03     0.0006    1e+03        1   ++
    3      5.3e+03    3.1e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 6/100
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000196
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.66          -0.088          -0.015              -1            -0.3               1           -0.35           -0.27          -0.022      5.6e+03      0.057       10        1   ++
    1           -0.66          -0.088          -0.015              -1            -0.3               1           -0.35           -0.27          -0.022      5.6e+03      0.057      4.5 -2.9e+05    -
    2           -0.66          -0.088          -0.015              -1            -0.3               1           -0.35           -0.27          -0.022      5.6e+03      0.057      2.2 -1.3e+02    -
    3           -0.66          -0.088          -0.015              -1            -0.3               1           -0.35           -0.27          -0.022      5.6e+03      0.057      1.1     -4.7    -
    4            -1.1            0.83            0.97            -1.6           -0.86             2.1          0.0079           -0.16           -0.75      5.6e+03      0.089      1.1     0.13    +
    5            -1.1            0.83            0.97            -1.6           -0.86             2.1          0.0079           -0.16           -0.75      5.6e+03      0.089     0.56     -1.2    -
    6            -1.1             1.1            0.92            -1.6            -0.3             2.1            0.12          -0.074           -0.72      5.4e+03      0.012     0.56     0.89    +
    7            -1.3             1.1            0.91            -1.7           -0.56             1.5          0.0027           -0.18            -0.7      5.4e+03      0.031      5.6     0.96   ++
    8            -1.3             1.1            0.91            -1.7           -0.56             1.5          0.0027           -0.18            -0.7      5.4e+03      0.031      1.1     -1.5    -
    9            -1.3             1.1            0.92            -1.6            -0.9            0.41            0.02           0.017           -0.68      5.3e+03     0.0093       11     0.95   ++
   10            -1.4             1.2               1            -1.5            -1.1            0.41          -0.084          -0.068           -0.65      5.3e+03    0.00049  1.1e+02        1   ++
   11            -1.4             1.2               1            -1.5            -1.1            0.41          -0.084          -0.068           -0.65      5.3e+03    5.2e-06  1.1e+02        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000197
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0              -1            0.26           0.015            -0.7           -0.83               1            -0.4           -0.16          -0.087      5.4e+03      0.043       10      1.1   ++
    1            -1.4             0.9            0.74            -1.1            -1.3             0.3           -0.34          -0.016           -0.48      5.3e+03      0.011    1e+02     0.98   ++
    2            -1.6             1.1            0.94            -1.1            -1.1            0.52           -0.28          -0.028           -0.55      5.3e+03     0.0015    1e+03      1.1   ++
    3            -1.6             1.2            0.97            -1.1            -1.1             0.6           -0.27          -0.028           -0.54      5.3e+03    0.00012    1e+04        1   ++
    4            -1.6             1.2            0.97            -1.1            -1.1             0.6           -0.27          -0.028           -0.54      5.3e+03    3.6e-07    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000198
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0              -1             0.4            0.04           -0.65               0               0               0               0               0           -0.29          -0.071           -0.12               0               0      5.6e+03      0.037       10      1.1   ++
    1            -1.2            0.89            0.71            -1.1               0               0               0               0               0           0.045          -0.044           -0.64               0               0      5.5e+03     0.0076    1e+02      1.1   ++
    2            -1.4             1.1            0.89            -1.1               0               0               0               0               0           0.057          -0.046           -0.68               0               0      5.5e+03    0.00067    1e+03      1.1   ++
    3            -1.4             1.1            0.91            -1.1               0               0               0               0               0           0.058          -0.046           -0.69               0               0      5.5e+03    6.1e-06    1e+04        1   ++
    4            -1.4             1.1            0.91            -1.1               0               0               0               0               0           0.058          -0.046           -0.69               0               0      5.5e+03    5.2e-10    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000199
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost     lambda_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.61            0.03          -0.014              -1             1.5           -0.47               1           -0.44           -0.34           -0.03      5.7e+03      0.076        1     0.86    +
    1            -1.1               1            0.08            -1.8            0.52            -1.2             1.2          -0.076          -0.037           -0.16      5.3e+03      0.019       10      0.9   ++
    2            -1.4             1.1             1.2            -1.6            0.46            -1.2            0.73            -0.1          -0.063           -0.55      5.3e+03     0.0066    1e+02      1.1   ++
    3            -1.4             1.2            0.97            -1.6            0.44            -1.1            0.58          -0.089          -0.064           -0.62      5.3e+03    0.00077    1e+03      1.1   ++
    4            -1.4             1.2            0.95            -1.5            0.44            -1.1            0.55          -0.087          -0.065           -0.63      5.3e+03    3.6e-05    1e+04        1   ++
    5            -1.4             1.2            0.95            -1.5            0.44            -1.1            0.55          -0.087          -0.065           -0.63      5.3e+03    3.4e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000200
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.69            0.53              -1             1.8               0               0               0               0               0           -0.57           -0.28               0               0        6e+03        0.1        1      0.6    +
    1            -1.1             1.5           -0.78             1.5               0               0               0               0               0            0.15           -0.67               0               0      5.4e+03      0.034       10     0.95   ++
    2            -1.1             1.5           -0.78             1.5               0               0               0               0               0            0.15           -0.67               0               0      5.4e+03      0.034      1.3     -3.9    -
    3           -0.93             2.1            -1.8            0.15               0               0               0               0               0            0.33            -1.1               0               0      5.2e+03      0.017      1.3     0.82    +
    4           -0.84             1.8            -1.6            0.31               0               0               0               0               0            0.25            -1.3               0               0      5.2e+03     0.0014       13        1   ++
    5           -0.87             1.8            -1.6            0.34               0               0               0               0               0            0.24            -1.4               0               0      5.2e+03    3.9e-05  1.3e+02        1   ++
    6           -0.87             1.8            -1.6            0.34               0               0               0               0               0            0.24            -1.4               0               0      5.2e+03    2.2e-08  1.3e+02        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000201
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.53             0.2              -1           -0.23               1         -0.0043           -0.53           -0.13      5.5e+03        2.5       10        1   ++
    1           -0.53             0.2              -1           -0.23               1         -0.0043           -0.53           -0.13      5.5e+03        2.5      4.5   -3e+05    -
    2           -0.53             0.2              -1           -0.23               1         -0.0043           -0.53           -0.13      5.5e+03        2.5      2.2 -1.1e+02    -
    3           -0.53             0.2              -1           -0.23               1         -0.0043           -0.53           -0.13      5.5e+03        2.5      1.1     -3.7    -
    4           -0.75             1.3            -1.6            -1.2             1.1         -0.0031           -0.22           -0.49      5.1e+03       0.22       11        1   ++
    5           -0.75             1.3            -1.6            -1.2             1.1         -0.0031           -0.22           -0.49      5.1e+03       0.22      1.1     -6.5    -
    6           -0.73             2.4            -1.7            -1.4            0.48          -0.007           -0.27           -0.99        5e+03       0.27       11        1   ++
    7           -0.94             2.2            -1.6            -1.7           -0.12         -0.0061           -0.31            -1.8      4.9e+03      0.021       11     0.79    +
    8           -0.93             2.2            -1.6            -1.5          -0.069         -0.0061           -0.28            -1.8      4.9e+03    0.00089  1.1e+02        1   ++
    9           -0.92             2.2            -1.6            -1.5          -0.038         -0.0061           -0.28            -1.9      4.9e+03    4.3e-05  1.1e+03        1   ++
   10           -0.92             2.2            -1.6            -1.5          -0.038         -0.0061           -0.28            -1.9      4.9e+03    3.9e-07  1.1e+03        1   ++
Considering neighbor 5/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:log [9880.76215679441, np.float64(9935.321844004699), 8]
Attempt 7/100
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000202
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0            -0.9            0.16          -0.013            0.76           -0.85              -1               1           -0.26           -0.21           -0.05           -0.33      5.2e+03      0.046       10      1.1   ++
    1            -1.7            0.52            0.36             1.8              -1            -1.9           -0.33           -0.52            0.13           -0.15            -1.5      5.1e+03      0.052       10     0.72    +
    2            -1.9            0.66             0.5             1.9           -0.97            -1.3           -0.27           -0.41            0.11           -0.47            -1.8        5e+03     0.0033    1e+02        1   ++
    3            -1.9            0.71            0.57             1.9              -1            -1.6            0.12           -0.44            0.12            -0.4            -1.9        5e+03     0.0057    1e+02     0.87    +
    4            -1.9            0.71            0.57             1.9              -1            -1.5            0.11           -0.42            0.11           -0.39            -1.8        5e+03    7.3e-05    1e+03     0.99   ++
    5            -1.9            0.71            0.57             1.9              -1            -1.5            0.11           -0.42            0.11           -0.39            -1.8        5e+03    1.8e-07    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000203
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0            -0.9               1           -0.68               0               0               0               0               0           -0.38           -0.45               0               0      5.5e+03      0.041       10      1.1   ++
    1              -1             1.7            -1.1               0               0               0               0               0           0.053            -1.2               0               0      5.3e+03     0.0097    1e+02      1.1   ++
    2            -1.1             1.8            -1.1               0               0               0               0               0           0.057            -1.4               0               0      5.3e+03    0.00044    1e+03        1   ++
    3            -1.1             1.8            -1.1               0               0               0               0               0           0.057            -1.4               0               0      5.3e+03    1.2e-06    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000204
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03      0.059       10        1   ++
    1      5.2e+03      0.025    1e+02     0.97   ++
    2      5.2e+03     0.0014    1e+03        1   ++
    3      5.2e+03    3.2e-05    1e+04        1   ++
    4      5.2e+03    1.6e-08    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000205
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      6.1e+03          2        1     0.51    +
    1      5.7e+03       0.32       10     0.92   ++
    2      5.7e+03       0.32        5 -2.1e+03    -
    3      5.7e+03       0.32      2.5      -18    -
    4      5.7e+03       0.32      1.2    -0.14    -
    5      5.5e+03       0.16      1.2     0.86    +
    6      5.4e+03      0.053       12        1   ++
    7      5.4e+03    0.00091  1.2e+02     0.97   ++
    8      5.4e+03    2.4e-06  1.2e+02        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000206
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0              -1            0.37           0.049           -0.63           -0.85             1.1         -0.0029            -0.6          -0.071           -0.13      5.4e+03        2.3       10      1.1   ++
    1            -1.2            0.92            0.75            -1.1            -1.3            0.23          -0.005           -0.45          -0.019           -0.51      5.3e+03       0.45    1e+02     0.93   ++
    2            -1.4             1.1            0.94            -1.1            -1.1            0.47         -0.0055           -0.39          -0.028           -0.56      5.3e+03      0.045    1e+03      1.1   ++
    3            -1.4             1.2            0.98            -1.1            -1.1            0.59         -0.0056           -0.38          -0.025           -0.54      5.3e+03    0.00073    1e+04        1   ++
    4            -1.4             1.2            0.97            -1.1            -1.1             0.6         -0.0056           -0.38          -0.026           -0.54      5.3e+03    7.7e-06    1e+05        1   ++
    5            -1.4             1.2            0.97            -1.1            -1.1             0.6         -0.0056           -0.38          -0.026           -0.54      5.3e+03    1.3e-06    1e+05        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000207
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.83               1           -0.71               0               0               0        -0.00098               0               0           -0.26           -0.43               0               0      5.4e+03        2.9       10      1.1   ++
    1           -0.83             1.7            -1.1               0               0               0         -0.0046               0               0          -0.054            -1.1               0               0      5.3e+03       0.58    1e+02      1.1   ++
    2           -0.85             1.8            -1.1               0               0               0          -0.006               0               0          -0.065            -1.3               0               0      5.3e+03      0.035    1e+03        1   ++
    3           -0.86             1.8            -1.1               0               0               0         -0.0061               0               0          -0.066            -1.3               0               0      5.3e+03    0.00014    1e+04        1   ++
    4           -0.86             1.8            -1.1               0               0               0         -0.0061               0               0          -0.066            -1.3               0               0      5.3e+03    2.2e-09    1e+04        1   ++
Considering neighbor 5/20 for current solution
Considering neighbor 6/20 for current solution
Attempt 8/100
Considering neighbor 0/20 for current solution
Attempt 9/100
Considering neighbor 0/20 for current solution
Attempt 10/100
Considering neighbor 0/20 for current solution
Attempt 11/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000208
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03      0.055       10        1   ++
    1      5.2e+03      0.027    1e+02     0.99   ++
    2      5.2e+03     0.0017    1e+03      1.1   ++
    3      5.2e+03    4.1e-05    1e+04        1   ++
    4      5.2e+03    2.5e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000209
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1            0.12          -0.012            0.86           -0.75           -0.92           -0.34            -0.1           -0.05           -0.22      5.2e+03      0.047       10      1.1   ++
    1            -1.5            0.49            0.36             1.7            -1.1            -1.1           -0.32            0.08           -0.21           -0.28        5e+03      0.016    1e+02      1.1   ++
    2            -1.7            0.68            0.56             1.8            -1.2            -1.1           -0.29           0.075           -0.25           -0.29        5e+03     0.0014    1e+03      1.1   ++
    3            -1.7            0.71            0.59             1.8            -1.2            -1.1           -0.29           0.074           -0.25           -0.29        5e+03    1.8e-05    1e+04        1   ++
    4            -1.7            0.71            0.59             1.8            -1.2            -1.1           -0.29           0.074           -0.25           -0.29        5e+03    3.4e-09    1e+04        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 12/100
Biogeme parameters read from biogeme.toml.
Model with 4 unknown parameters [max: 50]
*** Estimate b07everything_000210
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0              -1              -1           -0.38           -0.11      5.6e+03      0.041       10      1.1   ++
    1           -0.65            -2.8           -0.89          -0.037      5.3e+03      0.016    1e+02      1.1   ++
    2           -0.49            -3.3            -1.1         -0.0039      5.3e+03     0.0015    1e+03      1.1   ++
    3           -0.48            -3.4            -1.1         -0.0026      5.3e+03    9.8e-06    1e+04        1   ++
    4           -0.48            -3.4            -1.1         -0.0026      5.3e+03    4.4e-10    1e+04        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:no_seg;train_cost_catalog:linear;train_headway_catalog:without_headway;train_tt_catalog:sqrt [10592.228471637549, np.float64(10619.508315242692), 4]
Attempt 13/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000211
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.73            0.39              -1             1.6           -0.73           -0.45           -0.21      5.5e+03      0.049        1     0.83    +
    1            -1.1             1.4            -1.3             0.9            -1.1            -0.2           -0.35      5.1e+03      0.022       10      1.1   ++
    2           -0.87               2            -1.9            0.26            -1.1           0.037           -0.36        5e+03      0.012       10     0.82    +
    3              -1               2            -1.7            0.36            -1.1          -0.061           -0.31        5e+03     0.0012    1e+02        1   ++
    4              -1               2            -1.7            0.38            -1.1          -0.064           -0.31        5e+03    3.1e-05    1e+03        1   ++
    5              -1               2            -1.7            0.38            -1.1          -0.064           -0.31        5e+03    3.6e-09    1e+03        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:linear;train_headway_catalog:without_headway;train_tt_catalog:boxcox [10005.510774519118, np.float64(10053.25050082812), 7]
Attempt 14/100
Considering neighbor 0/20 for current solution
Attempt 15/100
Considering neighbor 0/20 for current solution
Attempt 16/100
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000212
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.73            0.22              -1           -0.27           -0.29           -0.13      5.5e+03      0.039       10        1   ++
    1            -1.1             2.3            -1.5            -2.5           -0.11             1.3        5e+03      0.028    1e+02     0.96   ++
    2            -1.1             2.2            -1.6            -2.8           -0.13             1.2        5e+03    0.00085    1e+03        1   ++
    3            -1.1             2.2            -1.6            -2.8           -0.13             1.2        5e+03    2.8e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:log [9958.803287791457, np.float64(9999.723053199174), 6]
Attempt 17/100
Biogeme parameters read from biogeme.toml.
Model with 22 unknown parameters [max: 50]
*** Estimate b07everything_000213
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03      0.039       10        1   ++
    1      5.6e+03     0.0083    1e+02      1.1   ++
    2      5.6e+03    0.00067    1e+03        1   ++
    3      5.6e+03    5.4e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 18/100
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000214
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03        2.7       10        1   ++
    1      5.2e+03       0.77    1e+02        1   ++
    2      5.2e+03      0.072    1e+03      1.1   ++
    3      5.2e+03     0.0013    1e+04        1   ++
    4      5.2e+03    5.2e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000215
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03        2.9        1     0.66    +
    1      5.7e+03       0.71        1     0.43    +
    2      5.7e+03       0.71      0.5     -1.9    -
    3      5.7e+03       0.71     0.25    -0.21    -
    4      5.7e+03      0.065     0.25     0.31    +
    5      5.6e+03      0.012      2.5      1.1   ++
    6      5.6e+03      0.012      1.2   -0.013    -
    7      5.4e+03       0.27      1.2     0.41    +
    8      5.3e+03      0.014       12        1   ++
    9      5.3e+03      0.011       12     0.74    +
   10      5.3e+03    0.00051  1.2e+02        1   ++
   11      5.3e+03    2.2e-06  1.2e+02        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 19/100
Considering neighbor 0/20 for current solution
Attempt 20/100
Biogeme parameters read from biogeme.toml.
Model with 22 unknown parameters [max: 50]
*** Estimate b07everything_000216
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03      0.041       10        1   ++
    1      5.5e+03      0.053    1e+02     0.97   ++
    2      5.5e+03     0.0043    1e+03        1   ++
    3      5.5e+03    0.00016    1e+04        1   ++
    4      5.5e+03    2.1e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000217
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.22            -0.6              -1               0               0               0         -0.0001               0               0           -0.25          0.0035               0               0      5.5e+03        2.7       10        1   ++
    1            0.67            -1.2            -1.5               0               0               0         -0.0046               0               0           -0.15            0.27               0               0      5.4e+03       0.46    1e+02        1   ++
    2            0.72            -1.3            -1.6               0               0               0         -0.0056               0               0           -0.17            0.27               0               0      5.4e+03      0.021    1e+03        1   ++
    3            0.72            -1.3            -1.6               0               0               0         -0.0057               0               0           -0.17            0.27               0               0      5.4e+03    4.4e-05    1e+04        1   ++
    4            0.72            -1.3            -1.6               0               0               0         -0.0057               0               0           -0.17            0.27               0               0      5.4e+03      2e-10    1e+04        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 21/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000218
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.91            0.18          -0.013             0.8           -0.85              -1           -0.25           -0.22          -0.054           -0.36      5.2e+03      0.044       10      1.1   ++
    1            -1.6            0.51            0.34             1.8           -0.98            -1.4           -0.42            0.12           -0.22            -1.4        5e+03       0.02    1e+02      1.1   ++
    2            -1.9            0.68            0.53             1.9              -1            -1.5           -0.42            0.11            -0.4            -1.8        5e+03     0.0015    1e+03      1.1   ++
    3            -1.9            0.71            0.56             1.9              -1            -1.5           -0.42            0.11           -0.42            -1.9        5e+03    1.7e-05    1e+04        1   ++
    4            -1.9            0.71            0.56             1.9              -1            -1.5           -0.42            0.11           -0.42            -1.9        5e+03    3.3e-09    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 22/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000219
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.71            0.47              -1             1.7           -0.56           -0.51           -0.24      5.7e+03       0.08        1      0.7    +
    1            -1.1             1.5              -1             1.3            -1.6            0.04           -0.42      5.1e+03      0.025       10        1   ++
    2            -1.1             1.5              -1             1.3            -1.6            0.04           -0.42      5.1e+03      0.025     0.81    -0.21    -
    3            -1.2             1.8            -1.5            0.44              -2          -0.059           -0.32        5e+03      0.018      8.1      1.1   ++
    4            -1.1             2.1            -1.6            0.33            -2.7           -0.12             1.3        5e+03     0.0031       81     0.98   ++
    5            -1.1             2.1            -1.6            0.34            -2.8           -0.12             1.2        5e+03    0.00014  8.1e+02        1   ++
    6            -1.1             2.1            -1.6            0.34            -2.8           -0.12             1.2        5e+03    5.9e-07  8.1e+02        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:boxcox [9922.718145551291, np.float64(9970.457871860293), 7]
Attempt 23/100
Considering neighbor 0/20 for current solution
Attempt 24/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000220
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.72            0.11         -0.0034           -0.97               2            -0.8         -0.0012           -0.29           -0.15           -0.05      5.9e+03        1.9        1     0.55    +
    1           -0.93            0.75           0.059           -0.68             1.7            -1.8          -0.008           -0.26          0.0067           -0.11      5.4e+03      0.021       10        1   ++
    2           -0.93            0.75           0.059           -0.68             1.7            -1.8          -0.008           -0.26          0.0067           -0.11      5.4e+03      0.021      4.1 -6.6e+02    -
    3           -0.93            0.75           0.059           -0.68             1.7            -1.8          -0.008           -0.26          0.0067           -0.11      5.4e+03      0.021        2      -12    -
    4           -0.93            0.75           0.059           -0.68             1.7            -1.8          -0.008           -0.26          0.0067           -0.11      5.4e+03      0.021        1    0.069    -
    5            -1.3            0.83            0.16            -1.5            0.72            -2.1        -0.00037           -0.24            0.33           -0.15      5.2e+03       0.19       10     0.97   ++
    6              -1             1.1             1.1            -1.7            0.43            -2.3         -0.0057           -0.14            0.16           -0.31      5.1e+03       0.12    1e+02     0.93   ++
    7            -1.1             1.2            0.95            -1.6            0.46            -2.4         -0.0055           -0.17            0.16           -0.23      5.1e+03     0.0019    1e+03        1   ++
    8            -1.1             1.2            0.95            -1.6            0.46            -2.4         -0.0055           -0.17            0.16           -0.23      5.1e+03    1.6e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 5 unknown parameters [max: 50]
*** Estimate b07everything_000221
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time       lambda_tt          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.68              -1             1.9           -0.64           -0.59        6e+03       0.11        1     0.59    +
    1           -0.92           -0.75             1.6            -1.6          -0.024      5.4e+03      0.031       10     0.96   ++
    2           -0.92           -0.75             1.6            -1.6          -0.024      5.4e+03      0.031      1.2     -1.7    -
    3           -0.68            -1.8            0.39            -2.3            0.03      5.3e+03       0.04      1.2     0.88    +
    4           -0.48            -1.7            0.47            -2.3           0.054      5.2e+03     0.0029       12     0.95   ++
    5            -0.5            -1.7            0.48            -2.4           0.057      5.2e+03    2.1e-05  1.2e+02        1   ++
    6            -0.5            -1.7            0.48            -2.4           0.057      5.2e+03    1.2e-09  1.2e+02        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:no_seg;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:boxcox [10501.31230191711, np.float64(10535.41210642354), 5]
Attempt 25/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000222
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC         asc_car beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.78           -0.65               0               0               0               0               0           -0.35               0               0      5.7e+03      0.035       10      1.1   ++
    1           -0.64            -1.1               0               0               0               0               0           0.032               0               0      5.6e+03     0.0056    1e+02        1   ++
    2           -0.63            -1.1               0               0               0               0               0           0.044               0               0      5.6e+03    8.2e-05    1e+03        1   ++
    3           -0.63            -1.1               0               0               0               0               0           0.044               0               0      5.6e+03    1.9e-08    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000223
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.97            0.51            0.33           -0.63               0               0               0         -0.0026               0               0           -0.47             0.1           -0.43               0               0      5.6e+03        2.7       10      1.1   ++
    1              -1            0.91            0.71            -1.1               0               0               0         -0.0047               0               0          -0.042          -0.054           -0.68               0               0      5.5e+03       0.35    1e+02      1.1   ++
    2            -1.2             1.1            0.89            -1.1               0               0               0         -0.0053               0               0           -0.05          -0.044           -0.68               0               0      5.5e+03      0.026    1e+03        1   ++
    3            -1.2             1.1            0.91            -1.1               0               0               0         -0.0054               0               0           -0.05          -0.044           -0.68               0               0      5.5e+03    0.00018    1e+04        1   ++
    4            -1.2             1.1            0.91            -1.1               0               0               0         -0.0054               0               0           -0.05          -0.044           -0.68               0               0      5.5e+03    1.1e-08    1e+04        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 26/100
Biogeme parameters read from biogeme.toml.
Model with 5 unknown parameters [max: 50]
*** Estimate b07everything_000224
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time       lambda_tt          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.69              -1             1.8           -0.64           -0.56      5.8e+03      0.086        1     0.66    +
    1           -0.41            -1.8            0.78            -1.6           -0.11      5.5e+03      0.034        1     0.87    +
    2           -0.65            -1.4             0.6           -0.95           -0.11      5.4e+03     0.0035       10     0.98   ++
    3           -0.59            -1.5            0.39              -1          -0.097      5.4e+03     0.0015    1e+02     0.95   ++
    4           -0.61            -1.5            0.41              -1           -0.11      5.4e+03      2e-05    1e+03        1   ++
    5           -0.61            -1.5            0.41              -1           -0.11      5.4e+03    8.8e-09    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 4 unknown parameters [max: 50]
*** Estimate b07everything_000225
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.79           -0.72              -1           -0.42      5.5e+03      0.037       10      1.1   ++
    1           -0.75            -1.2            -2.2            -0.1      5.3e+03     0.0076    1e+02      1.1   ++
    2           -0.73            -1.2            -2.3           -0.11      5.3e+03    0.00023    1e+03        1   ++
    3           -0.73            -1.2            -2.3           -0.11      5.3e+03    2.7e-07    1e+03        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:no_seg;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:linear [10585.801045811912, np.float64(10613.080889417055), 4]
Attempt 27/100
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000226
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.53          -0.053         -0.0067            0.32              -1           -0.41               1          -0.004           -0.39           -0.22          -0.028           -0.15      5.4e+03        2.4       10        1   ++
    1           -0.53          -0.053         -0.0067            0.32              -1           -0.41               1          -0.004           -0.39           -0.22          -0.028           -0.15      5.4e+03        2.4      1.8      -24    -
    2           -0.97            0.41            0.11             2.2            -1.6           -0.96             0.7          -0.004           -0.11           0.036           -0.15           -0.86        5e+03        1.1       18        1   ++
    3           -0.97            0.41            0.11             2.2            -1.6           -0.96             0.7          -0.004           -0.11           0.036           -0.15           -0.86        5e+03        1.1      1.1      -13    -
    4            -1.2            0.68            0.16               2            -1.7            -1.7           -0.36         -0.0061           -0.26           -0.04           -0.25            -1.2        5e+03       0.13      1.1     0.26    +
    5            -1.4             0.7             0.4             2.1            -1.6            -1.5           -0.24         -0.0061           -0.31           0.049           -0.45            -1.7      4.9e+03     0.0045       11      1.1   ++
    6            -1.4            0.71            0.42             2.1            -1.6            -1.5          -0.054         -0.0061           -0.32           0.048           -0.46            -1.8      4.9e+03     0.0019  1.1e+02      1.1   ++
    7            -1.4            0.71            0.42             2.1            -1.6            -1.5          -0.034         -0.0061           -0.31           0.045           -0.45            -1.8      4.9e+03    6.4e-05  1.1e+03        1   ++
    8            -1.4            0.71            0.42             2.1            -1.6            -1.5          -0.034         -0.0061           -0.31           0.045           -0.45            -1.8      4.9e+03    3.2e-07  1.1e+03        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:LUGGAGE-GA;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:log [9830.542339359698, np.float64(9912.38187017513), 12]
Attempt 28/100
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000227
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1           -0.33          -0.026               0               0               0           -0.22               0               0           -0.38           -0.27           -0.02               0               0      5.9e+03       0.04       10        1   ++
    1            -2.1             0.8            0.74               0               0               0            -1.8               0               0           -0.85            0.28            0.22               0               0      5.5e+03      0.021    1e+02      1.1   ++
    2            -2.5             1.2             1.1               0               0               0            -2.2               0               0            -0.9            0.24           -0.01               0               0      5.5e+03     0.0037    1e+03      1.1   ++
    3            -2.6             1.2             1.2               0               0               0            -2.2               0               0           -0.91            0.23          -0.038               0               0      5.5e+03    0.00012    1e+04        1   ++
    4            -2.6             1.2             1.2               0               0               0            -2.2               0               0           -0.91            0.23          -0.038               0               0      5.5e+03    1.1e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000228
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1           -0.15          -0.015               0               0               0           -0.53         -0.0043               0               0           -0.67           -0.32          -0.045               0               0      5.8e+03        2.5       10        1   ++
    1            -1.9            0.85            0.79               0               0               0              -2         -0.0051               0               0           -0.97            0.27            0.16               0               0      5.5e+03       0.35    1e+02      1.1   ++
    2            -2.2             1.2             1.1               0               0               0            -2.2         -0.0055               0               0              -1            0.24          -0.022               0               0      5.5e+03      0.063    1e+03      1.1   ++
    3            -2.3             1.2             1.2               0               0               0            -2.2         -0.0055               0               0              -1            0.24          -0.035               0               0      5.5e+03     0.0015    1e+04        1   ++
    4            -2.3             1.2             1.2               0               0               0            -2.2         -0.0055               0               0              -1            0.24          -0.035               0               0      5.5e+03    8.7e-07    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000229
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1           -0.17          -0.015               0               0               0           -0.53         -0.0047               0               0           -0.73           -0.37          -0.047               0               0      5.8e+03        2.4       10        1   ++
    1            -1.9            0.83            0.76               0               0               0            -1.1         -0.0052               0               0              -1            0.05           -0.39               0               0      5.5e+03       0.53    1e+02      1.1   ++
    2            -2.3             1.2             1.1               0               0               0            -1.1         -0.0055               0               0              -1           0.028           -0.47               0               0      5.5e+03      0.073    1e+03      1.1   ++
    3            -2.3             1.2             1.2               0               0               0            -1.1         -0.0055               0               0              -1           0.028           -0.48               0               0      5.5e+03     0.0018    1e+04        1   ++
    4            -2.3             1.2             1.2               0               0               0            -1.1         -0.0055               0               0              -1           0.028           -0.48               0               0      5.5e+03    1.2e-06    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000230
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.61           0.037          -0.014              -1             1.5           -0.44           -0.45           -0.34           -0.03      5.7e+03       0.08        1     0.84    +
    1            -1.1               1            0.12            -1.8            0.54            -1.2          -0.022           0.043           -0.23      5.3e+03      0.012       10     0.95   ++
    2            -1.4             1.1             1.1            -1.5            0.43            -1.1          -0.081          -0.081           -0.67      5.3e+03     0.0015    1e+02     0.98   ++
    3            -1.4             1.1            0.95            -1.5            0.38            -1.1          -0.081          -0.077           -0.73      5.3e+03    8.8e-05    1e+03        1   ++
    4            -1.4             1.1            0.95            -1.5            0.38            -1.1          -0.081          -0.077           -0.73      5.3e+03    1.1e-07    1e+03        1   ++
Considering neighbor 3/20 for current solution
Considering neighbor 4/20 for current solution
Attempt 29/100
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000231
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03        2.5       10        1   ++
    1      5.7e+03        2.5      3.3 -1.1e+04    -
    2      5.7e+03        2.5      1.7      -13    -
    3      5.3e+03        1.2       17      1.1   ++
    4      5.3e+03        1.2      1.2      -29    -
    5      5.3e+03        1.2     0.62   -0.081    -
    6      5.2e+03      0.011      6.2     0.98   ++
    7      5.2e+03     0.0076       62     0.96   ++
    8      5.2e+03     0.0018  6.2e+02      1.1   ++
    9      5.2e+03    1.2e-05  6.2e+03        1   ++
   10      5.2e+03    1.1e-09  6.2e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000232
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1          -0.065               0               0               0           -0.27               0               0          -0.047           -0.85               0               0        6e+03      0.075       10      1.1   ++
    1              -2             2.2               0               0               0           -0.88               0               0           -0.84           0.038               0               0      5.3e+03      0.038    1e+02        1   ++
    2            -2.2             2.1               0               0               0              -1               0               0           -0.88          -0.067               0               0      5.3e+03     0.0012    1e+03        1   ++
    3            -2.2             2.1               0               0               0              -1               0               0           -0.88          -0.067               0               0      5.3e+03    5.4e-06    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000233
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91               0               0               0            -0.3         -0.0074               0               0              -1               0               0      5.9e+03        2.6       10        1   ++
    1            -1.4               0               0               0              -2         -0.0057               0               0           -0.78               0               0      5.6e+03       0.22    1e+02     0.99   ++
    2            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.87               0               0      5.6e+03     0.0089    1e+03        1   ++
    3            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.88               0               0      5.6e+03    1.2e-05    1e+04        1   ++
    4            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.88               0               0      5.6e+03    4.4e-11    1e+04        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 30/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000234
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.67            0.14          -0.011              -1             1.8               0               0               0               0               0           -0.37            -0.3           -0.04               0               0        6e+03      0.092        1      0.6    +
    1            -1.3             1.1            0.11            -1.3             1.1               0               0               0               0               0            0.15          -0.018           -0.23               0               0      5.5e+03      0.018       10      1.1   ++
    2           -0.74             1.1             1.1            -2.2             0.1               0               0               0               0               0            0.56           -0.13           -0.84               0               0      5.5e+03      0.042       10     0.29    +
    3            -1.2             1.1             0.9            -1.6            0.28               0               0               0               0               0            0.24          -0.079           -0.78               0               0      5.5e+03     0.0026    1e+02        1   ++
    4            -1.2             1.1            0.88            -1.6            0.43               0               0               0               0               0            0.25          -0.081           -0.78               0               0      5.5e+03     0.0013    1e+03     0.95   ++
    5            -1.2             1.1            0.89            -1.5            0.42               0               0               0               0               0            0.24          -0.079           -0.77               0               0      5.5e+03    7.2e-06    1e+04        1   ++
    6            -1.2             1.1            0.89            -1.5            0.42               0               0               0               0               0            0.24          -0.079           -0.77               0               0      5.5e+03    1.1e-09    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 31/100
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000235
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.5e+03        2.6       10        1   ++
    1      5.2e+03       0.87    1e+02        1   ++
    2      5.2e+03      0.069    1e+03      1.1   ++
    3      5.2e+03     0.0012    1e+04        1   ++
    4      5.2e+03    4.5e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000236
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03      0.059       10        1   ++
    1      5.2e+03      0.025    1e+02     0.97   ++
    2      5.2e+03     0.0014    1e+03        1   ++
    3      5.2e+03    3.2e-05    1e+04        1   ++
    4      5.2e+03    1.6e-08    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000237
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.22           -0.77           -0.62               0               0               0               0               0           -0.42           0.022               0               0      5.6e+03      0.038       10      1.1   ++
    1             0.1            -1.1            -1.1               0               0               0               0               0           -0.29            0.33               0               0      5.5e+03     0.0078    1e+02      1.1   ++
    2            0.13            -1.2            -1.1               0               0               0               0               0           -0.28            0.33               0               0      5.5e+03    0.00024    1e+03        1   ++
    3            0.13            -1.2            -1.1               0               0               0               0               0           -0.28            0.33               0               0      5.5e+03    3.2e-07    1e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000238
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.5e+03        2.6       10        1   ++
    1      5.2e+03       0.87    1e+02        1   ++
    2      5.2e+03      0.069    1e+03      1.1   ++
    3      5.2e+03     0.0012    1e+04        1   ++
    4      5.2e+03    4.5e-07    1e+04        1   ++
Considering neighbor 3/20 for current solution
Considering neighbor 4/20 for current solution
Attempt 32/100
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000239
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1           -0.33          -0.026               0               0               0           -0.19               0               0           -0.38           -0.27          -0.021               0               0      5.9e+03       0.04       10        1   ++
    1            -2.1             0.8            0.71               0               0               0              -1               0               0           -0.87           0.072           -0.35               0               0      5.6e+03      0.024    1e+02      1.1   ++
    2            -2.5             1.2             1.1               0               0               0            -1.1               0               0            -0.9           0.028           -0.48               0               0      5.5e+03      0.004    1e+03      1.1   ++
    3            -2.6             1.2             1.2               0               0               0            -1.1               0               0            -0.9           0.027           -0.48               0               0      5.5e+03    0.00013    1e+04        1   ++
    4            -2.6             1.2             1.2               0               0               0            -1.1               0               0            -0.9           0.027           -0.48               0               0      5.5e+03    1.2e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000240
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1            0.15               0               0               0           -0.35               0               0          -0.045              -1               0               0      5.9e+03      0.076       10      1.1   ++
    1            -2.1             2.2               0               0               0            -1.2               0               0           -0.86            -1.1               0               0      5.2e+03      0.031    1e+02        1   ++
    2            -2.3             2.2               0               0               0            -1.5               0               0           -0.95            -1.5               0               0      5.2e+03     0.0012    1e+03        1   ++
    3            -2.3             2.2               0               0               0            -1.5               0               0           -0.95            -1.5               0               0      5.2e+03    3.9e-06    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000241
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost     lambda_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91            0.23               0               0               0           -0.22               1         -0.0077               0               0              -1            -0.2               0               0      5.8e+03        2.6       10        1   ++
    1           -0.91            0.23               0               0               0           -0.22               1         -0.0077               0               0              -1            -0.2               0               0      5.8e+03        2.6      4.5 -2.4e+05    -
    2           -0.91            0.23               0               0               0           -0.22               1         -0.0077               0               0              -1            -0.2               0               0      5.8e+03        2.6      2.2      -98    -
    3           -0.91            0.23               0               0               0           -0.22               1         -0.0077               0               0              -1            -0.2               0               0      5.8e+03        2.6      1.1     -3.4    -
    4            -1.6             1.4               0               0               0            -1.1               1         -0.0053               0               0           -0.96           -0.53               0               0      5.4e+03       0.29       11        1   ++
    5            -1.6             1.4               0               0               0            -1.1               1         -0.0053               0               0           -0.96           -0.53               0               0      5.4e+03       0.29      3.2   -2e+04    -
    6            -1.6             1.4               0               0               0            -1.1               1         -0.0053               0               0           -0.96           -0.53               0               0      5.4e+03       0.29      1.6      -43    -
    7            -2.2               3               0               0               0              -2           -0.32         -0.0068               0               0            -1.3            -1.1               0               0      5.3e+03       0.44      1.6     0.46    +
    8              -2             2.2               0               0               0            -1.3           -0.31         -0.0062               0               0              -1            -1.6               0               0      5.2e+03      0.049       16     0.93   ++
    9              -2             2.3               0               0               0            -1.5           -0.13         -0.0064               0               0            -1.1            -1.7               0               0      5.2e+03     0.0049  1.6e+02     0.92   ++
   10              -2             2.3               0               0               0            -1.5           -0.15         -0.0064               0               0            -1.1            -1.7               0               0      5.2e+03    5.1e-05  1.6e+03        1   ++
   11              -2             2.3               0               0               0            -1.5           -0.15         -0.0064               0               0            -1.1            -1.7               0               0      5.2e+03    5.9e-09  1.6e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000242
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1               0               0               0           -0.38               0               0           -0.11               0               0        6e+03      0.074       10      1.1   ++
    1            -1.6               0               0               0           -0.83               0               0           -0.75               0               0      5.6e+03     0.0058    1e+02        1   ++
    2            -1.7               0               0               0           -0.94               0               0            -0.8               0               0      5.6e+03    0.00016    1e+03        1   ++
    3            -1.7               0               0               0           -0.94               0               0            -0.8               0               0      5.6e+03    1.4e-07    1e+03        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000243
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1               0               0               0           -0.28               0               0           -0.11               0               0        6e+03      0.073       10      1.1   ++
    1            -1.6               0               0               0           -0.94               0               0           -0.79               0               0      5.7e+03      0.006    1e+02      1.1   ++
    2            -1.7               0               0               0            -1.1               0               0           -0.87               0               0      5.7e+03    0.00022    1e+03        1   ++
    3            -1.7               0               0               0            -1.1               0               0           -0.87               0               0      5.7e+03      4e-07    1e+03        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000244
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1           -0.18          -0.017               0               0               0           -0.67          -0.005               0               0           -0.74           -0.36          -0.043               0               0      5.7e+03        2.3       10        1   ++
    1            -1.9            0.85             0.8               0               0               0           -0.94         -0.0052               0               0              -1            0.18          -0.054               0               0      5.5e+03       0.56    1e+02      1.1   ++
    2            -2.2             1.2             1.1               0               0               0           -0.96         -0.0055               0               0              -1            0.16          -0.097               0               0      5.5e+03      0.071    1e+03      1.1   ++
    3            -2.3             1.2             1.2               0               0               0           -0.96         -0.0056               0               0              -1            0.16          -0.097               0               0      5.5e+03     0.0017    1e+04        1   ++
    4            -2.3             1.2             1.2               0               0               0           -0.96         -0.0056               0               0              -1            0.16          -0.097               0               0      5.5e+03    1.1e-06    1e+04        1   ++
Considering neighbor 5/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000245
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost     lambda_cost beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08       10        1   ++
    1              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08        5 -8.3e+05    -
    2              -1               0               0               0            -0.3               2               0               0           -0.13               0               0        6e+03       0.08      2.5      -42    -
    3            -1.5               0               0               0            -1.7            -0.5               0               0            -1.3               0               0      5.9e+03      0.082      2.5      0.2    +
    4            -1.5               0               0               0            -1.7            -0.5               0               0            -1.3               0               0      5.9e+03      0.082        1    -0.66    -
    5            -1.4               0               0               0           -0.71           -0.45               0               0           -0.58               0               0      5.7e+03      0.026        1     0.82    +
    6            -1.7               0               0               0            -1.4            0.59               0               0           -0.93               0               0      5.7e+03      0.026        1     0.38    +
    7            -1.7               0               0               0            -1.1            0.38               0               0           -0.88               0               0      5.7e+03     0.0015       10      1.1   ++
    8            -1.7               0               0               0            -1.1            0.17               0               0           -0.88               0               0      5.7e+03     0.0012    1e+02     0.98   ++
    9            -1.7               0               0               0            -1.1            0.17               0               0           -0.88               0               0      5.7e+03    3.3e-06    1e+02        1   ++
Considering neighbor 6/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000246
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1               0               0               0           -0.38               0               0           -0.11               0               0        6e+03      0.074       10      1.1   ++
    1            -1.6               0               0               0           -0.83               0               0           -0.75               0               0      5.6e+03     0.0058    1e+02        1   ++
    2            -1.7               0               0               0           -0.94               0               0            -0.8               0               0      5.6e+03    0.00016    1e+03        1   ++
    3            -1.7               0               0               0           -0.94               0               0            -0.8               0               0      5.6e+03    1.4e-07    1e+03        1   ++
Considering neighbor 7/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000247
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91            0.23               0               0               0           -0.22         -0.0078               0               0              -1            -0.2               0               0      5.8e+03        2.6       10        1   ++
    1            -1.8             2.3               0               0               0            -1.4         -0.0061               0               0           -0.98            -1.2               0               0      5.2e+03       0.85    1e+02     0.99   ++
    2              -2             2.3               0               0               0            -1.5         -0.0063               0               0            -1.1            -1.6               0               0      5.2e+03      0.044    1e+03        1   ++
    3              -2             2.3               0               0               0            -1.5         -0.0064               0               0            -1.1            -1.6               0               0      5.2e+03    0.00035    1e+04        1   ++
    4              -2             2.3               0               0               0            -1.5         -0.0064               0               0            -1.1            -1.6               0               0      5.2e+03    1.5e-08    1e+04        1   ++
Considering neighbor 8/20 for current solution
Considering neighbor 9/20 for current solution
Attempt 33/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000248
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.47            -0.7            0.44              -1           -0.64               1          0.0022           -0.28          -0.067           -0.19      5.3e+03        2.7       10      1.1   ++
    1           -0.47            -0.7            0.44              -1           -0.64               1          0.0022           -0.28          -0.067           -0.19      5.3e+03        2.7      1.3     -3.2    -
    2           -0.12              -1             1.7            -1.4            -1.2            0.84         -0.0024           -0.33            0.21           -0.62      4.9e+03        0.7       13        1   ++
    3           -0.12              -1             1.7            -1.4            -1.2            0.84         -0.0024           -0.33            0.21           -0.62      4.9e+03        0.7     0.64    0.017    -
    4           0.019            -1.2             2.2            -1.6            -1.4             0.2         -0.0063           -0.52            0.31            -1.1      4.8e+03       0.15      6.4      1.1   ++
    5          -0.071            -1.2             2.2            -1.6            -1.5          -0.045         -0.0066           -0.68            0.45            -1.8      4.8e+03       0.03       64        1   ++
    6          -0.071            -1.2             2.2            -1.6            -1.5          -0.036         -0.0066           -0.69            0.46              -2      4.8e+03     0.0014  6.4e+02        1   ++
    7          -0.071            -1.2             2.2            -1.6            -1.5          -0.036         -0.0066           -0.69            0.46              -2      4.8e+03    4.6e-06  6.4e+02        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:MALE-GA;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:log [9640.412331069461, np.float64(9708.611940082323), 10]
Attempt 34/100
Considering neighbor 0/20 for current solution
Attempt 35/100
Considering neighbor 0/20 for current solution
Attempt 36/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000249
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.4e+03      0.045       10      1.1   ++
    1      5.3e+03      0.012    1e+02      1.1   ++
    2      5.3e+03     0.0009    1e+03        1   ++
    3      5.3e+03    7.2e-06    1e+04        1   ++
    4      5.3e+03    6.1e-10    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000250
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03       0.04       10      1.1   ++
    1      5.9e+03       0.04        5 -1.1e+07    -
    2      5.9e+03       0.04      2.5 -1.8e+02    -
    3      5.9e+03       0.04      1.2    0.051    -
    4      5.4e+03      0.043       12      1.1   ++
    5      5.4e+03      0.043      2.7 -3.1e+03    -
    6      5.4e+03      0.043      1.4      -17    -
    7      5.3e+03      0.057      1.4     0.48    +
    8      5.2e+03     0.0044       14     0.91   ++
    9      5.2e+03     0.0019       14     0.89    +
   10      5.2e+03    3.7e-05  1.4e+02        1   ++
   11      5.2e+03      5e-08  1.4e+02        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000251
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03      0.038       10        1   ++
    1      5.6e+03      0.038     0.95     -1.8    -
    2      5.4e+03      0.021      9.5     0.92   ++
    3      5.4e+03    0.00087       95     0.97   ++
    4      5.4e+03    2.2e-05  9.5e+02     0.99   ++
    5      5.4e+03    3.2e-08  9.5e+02        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000252
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1           -0.24          -0.026            0.25           -0.81           -0.28           -0.52           -0.37          -0.027           -0.17      5.7e+03      0.052       10        1   ++
    1            -1.3            0.35            0.16             2.2            -2.5            -1.3           -0.23            0.13           -0.18            -1.5        5e+03      0.032    1e+02     0.97   ++
    2            -1.6            0.63            0.43             1.9            -2.9            -1.5           -0.24           0.083           -0.43            -1.9      4.9e+03      0.002    1e+03        1   ++
    3            -1.7             0.7            0.49             1.9            -2.9            -1.5           -0.23           0.078           -0.46            -1.9      4.9e+03    6.6e-05    1e+04        1   ++
    4            -1.7             0.7            0.49             1.9            -2.9            -1.5           -0.23           0.078           -0.46            -1.9      4.9e+03    5.2e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000253
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.37           -0.65              -1           -0.83          0.0008            -0.3          -0.073      5.3e+03        2.4       10      1.1   ++
    1             0.6            -1.3            -1.6              -1         -0.0046           -0.37            0.25      5.2e+03       0.49    1e+02      1.1   ++
    2            0.68            -1.4            -1.7              -1         -0.0057           -0.36            0.25      5.2e+03      0.021    1e+03        1   ++
    3            0.68            -1.4            -1.7              -1         -0.0057           -0.35            0.24      5.2e+03    2.6e-05    1e+04        1   ++
    4            0.68            -1.4            -1.7              -1         -0.0057           -0.35            0.24      5.2e+03    1.1e-08    1e+04        1   ++
Considering neighbor 4/20 for current solution
Considering neighbor 5/20 for current solution
Attempt 37/100
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000254
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.83            0.17          0.0037               1           -0.76           -0.99               1         -0.0043           -0.48           -0.17          -0.071           -0.42      5.2e+03        2.5       10      1.1   ++
    1           -0.83            0.17          0.0037               1           -0.76           -0.99               1         -0.0043           -0.48           -0.17          -0.071           -0.42      5.2e+03        2.5     0.82    0.069    -
    2            -1.1            0.57           0.066             1.8            -1.2            -1.2             0.6         -0.0075            -0.4          -0.018           -0.14           -0.81        5e+03       0.69      8.2      1.1   ++
    3            -1.6            0.69            0.52             1.9              -1            -1.6          -0.016         -0.0064           -0.56            0.11           -0.33            -1.7        5e+03      0.033      8.2      0.9    +
    4            -1.6            0.71            0.55             1.9              -1            -1.5           0.081         -0.0063           -0.55            0.12           -0.39            -1.8        5e+03    0.00094       82      1.1   ++
    5            -1.6            0.71            0.55             1.9              -1            -1.5            0.11         -0.0063           -0.54            0.11           -0.39            -1.8        5e+03    4.2e-05  8.2e+02        1   ++
    6            -1.6            0.71            0.55             1.9              -1            -1.5            0.11         -0.0063           -0.54            0.11           -0.39            -1.8        5e+03    1.3e-07  8.2e+02        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 38/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000255
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.41           -0.74              -1               0               0               0        -0.00051               0               0           -0.44          -0.088               0               0      5.6e+03        2.8       10        1   ++
    1            0.54            -1.2            -2.8               0               0               0         -0.0045               0               0           -0.22             0.3               0               0      5.4e+03       0.37    1e+02      1.1   ++
    2            0.65            -1.2              -3               0               0               0         -0.0056               0               0           -0.23            0.32               0               0      5.4e+03      0.019    1e+03        1   ++
    3            0.65            -1.2              -3               0               0               0         -0.0057               0               0           -0.23            0.32               0               0      5.4e+03    4.8e-05    1e+04        1   ++
    4            0.65            -1.2              -3               0               0               0         -0.0057               0               0           -0.23            0.32               0               0      5.4e+03    2.5e-10    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 19 unknown parameters [max: 50]
*** Estimate b07everything_000256
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03        2.9       10        1   ++
    1      5.7e+03       0.32    1e+02      1.1   ++
    2      5.7e+03      0.014    1e+03        1   ++
    3      5.7e+03    2.1e-05    1e+04        1   ++
    4      5.7e+03    4.6e-11    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000257
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91            0.23               0               0               0            -0.3         -0.0079               0               0              -1            -0.2               0               0      5.8e+03        2.6       10        1   ++
    1            -1.8             2.3               0               0               0            -2.7         -0.0061               0               0           -0.98             1.5               0               0      5.2e+03       0.78    1e+02     0.97   ++
    2              -2             2.2               0               0               0            -2.7         -0.0064               0               0            -1.1             1.4               0               0      5.2e+03      0.046    1e+03        1   ++
    3              -2             2.2               0               0               0            -2.7         -0.0064               0               0            -1.1             1.4               0               0      5.2e+03    0.00027    1e+04        1   ++
    4              -2             2.2               0               0               0            -2.7         -0.0064               0               0            -1.1             1.4               0               0      5.2e+03    8.8e-09    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 19 unknown parameters [max: 50]
*** Estimate b07everything_000258
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03        2.7       10        1   ++
    1      5.5e+03       0.93    1e+02     0.97   ++
    2      5.5e+03      0.053    1e+03        1   ++
    3      5.5e+03    0.00034    1e+04        1   ++
    4      5.5e+03    1.4e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000259
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.34           -0.75           -0.64              -1         0.00013           -0.33          -0.036      5.3e+03        2.7       10      1.1   ++
    1            0.23            -1.2            -1.1            -2.1         -0.0045           -0.47            0.27      5.1e+03        0.4    1e+02      1.1   ++
    2            0.33            -1.2            -1.2            -2.3         -0.0057           -0.54            0.32      5.1e+03      0.019    1e+03        1   ++
    3            0.34            -1.2            -1.2            -2.4         -0.0058           -0.55            0.33      5.1e+03    5.5e-05    1e+04        1   ++
    4            0.34            -1.2            -1.2            -2.4         -0.0058           -0.55            0.33      5.1e+03      6e-08    1e+04        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000260
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.95            0.38              -1           -0.61         -0.0045           -0.85           -0.21      5.4e+03        2.4       10        1   ++
    1           -0.76             2.1            -2.9              -1         -0.0053           -0.22           -0.29        5e+03       0.91    1e+02     0.99   ++
    2           -0.77               2            -3.2            -1.1         -0.0061           -0.21           -0.29        5e+03      0.045    1e+03        1   ++
    3           -0.77               2            -3.2            -1.1         -0.0062           -0.21            -0.3        5e+03    0.00024    1e+04        1   ++
    4           -0.77               2            -3.2            -1.1         -0.0062           -0.21            -0.3        5e+03    9.4e-07    1e+04        1   ++
Considering neighbor 5/20 for current solution
Considering neighbor 6/20 for current solution
Attempt 39/100
Biogeme parameters read from biogeme.toml.
Model with 21 unknown parameters [max: 50]
*** Estimate b07everything_000261
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03        2.5       10        1   ++
    1      5.8e+03       0.56    1e+02      1.1   ++
    2      5.7e+03       0.07    1e+03      1.1   ++
    3      5.7e+03     0.0016    1e+04        1   ++
    4      5.7e+03    9.4e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000262
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03        2.5       10        1   ++
    1      5.2e+03          1    1e+02        1   ++
    2      5.2e+03      0.088    1e+03      1.1   ++
    3      5.2e+03     0.0024    1e+04        1   ++
    4      5.2e+03      2e-06    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000263
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0              -1          -0.065               0               0               0           -0.72               0               0          -0.048           -0.46               0               0      5.9e+03      0.076       10      1.1   ++
    1            -2.1             2.3               0               0               0            -2.3               0               0           -0.87             1.3               0               0      5.3e+03      0.038    1e+02        1   ++
    2            -2.3             2.2               0               0               0            -2.7               0               0           -0.93             1.4               0               0      5.3e+03     0.0012    1e+03        1   ++
    3            -2.3             2.2               0               0               0            -2.7               0               0           -0.93             1.4               0               0      5.3e+03    5.7e-06    1e+03        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 40/100
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000264
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0              -1            0.34            0.05           -0.62           -0.97          -0.003           -0.59          -0.089           -0.14      5.4e+03        2.3       10      1.1   ++
    1            -1.2            0.91             0.7           -0.99            -1.1         -0.0051            -0.4          -0.024           -0.61      5.3e+03       0.33    1e+02      1.1   ++
    2            -1.4             1.1            0.93              -1            -1.1         -0.0055           -0.39          -0.036           -0.66      5.3e+03      0.034    1e+03      1.1   ++
    3            -1.4             1.2            0.96              -1            -1.1         -0.0055           -0.39          -0.036           -0.65      5.3e+03    0.00019    1e+04        1   ++
    4            -1.4             1.2            0.96              -1            -1.1         -0.0055           -0.39          -0.036           -0.65      5.3e+03    3.1e-06    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000265
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost     lambda_cost       b_headway         asc_car     Function    Relgrad   Radius      Rho
    0            -0.8           -0.64           -0.82             1.1         -0.0026           -0.61      5.5e+03        2.2       10      1.1   ++
    1           -0.49            -1.1            -1.2            0.56         -0.0056           -0.41      5.4e+03       0.26    1e+02        1   ++
    2           -0.53            -1.2            -1.1            0.62         -0.0054           -0.37      5.4e+03     0.0047    1e+03        1   ++
    3           -0.53            -1.2            -1.1            0.62         -0.0054           -0.37      5.4e+03    4.7e-06    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000266
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.22           -0.78           -0.62           -0.87               2         -0.0029           -0.76           0.094      5.6e+03        2.2       10     0.94   ++
    1           -0.22           -0.78           -0.62           -0.87               2         -0.0029           -0.76           0.094      5.6e+03        2.2      4.2 -1.5e+05    -
    2           -0.22           -0.78           -0.62           -0.87               2         -0.0029           -0.76           0.094      5.6e+03        2.2      2.1 -1.3e+02    -
    3           -0.22           -0.78           -0.62           -0.87               2         -0.0029           -0.76           0.094      5.6e+03        2.2      1.1      -10    -
    4           -0.22           -0.78           -0.62           -0.87               2         -0.0029           -0.76           0.094      5.6e+03        2.2     0.53    -0.41    -
    5            0.11           -0.97            -1.1           -0.59             1.8         -0.0034           -0.59            0.36      5.3e+03       0.54      5.3      1.1   ++
    6            0.11           -0.97            -1.1           -0.59             1.8         -0.0034           -0.59            0.36      5.3e+03       0.54      2.6 -5.9e+02    -
    7            0.11           -0.97            -1.1           -0.59             1.8         -0.0034           -0.59            0.36      5.3e+03       0.54      1.3     -5.2    -
    8            0.56            -1.5            -1.2            -1.2            0.46         -0.0063           -0.66            0.32      5.2e+03       0.23      1.3     0.61    +
    9            0.28            -1.3            -1.1            -1.1            0.58         -0.0058           -0.71            0.32      5.2e+03      0.015       13        1   ++
   10            0.29            -1.2            -1.1            -1.1            0.61         -0.0059            -0.7            0.33      5.2e+03    5.9e-05  1.3e+02        1   ++
   11            0.29            -1.2            -1.1            -1.1            0.61         -0.0059            -0.7            0.33      5.2e+03      8e-08  1.3e+02        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 41/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000267
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.65           0.046         -0.0052           -0.98             1.8              -1         -0.0022           -0.35           -0.18          -0.042      5.6e+03        1.6        1     0.76    +
    1           -0.98               1           0.075            -1.7            0.81            -1.1         -0.0048           -0.34            0.15           -0.11      5.2e+03       0.56       10        1   ++
    2              -1             1.1             1.1            -1.7            0.53            -1.1         -0.0055           -0.13           0.061           -0.41      5.2e+03      0.029    1e+02     0.99   ++
    3            -1.1             1.2            0.96            -1.6             0.5            -1.1         -0.0055           -0.18           0.071            -0.3      5.2e+03     0.0026    1e+03        1   ++
    4            -1.1             1.2            0.96            -1.6             0.5            -1.1         -0.0055           -0.18           0.071            -0.3      5.2e+03    6.7e-06    1e+04        1   ++
    5            -1.1             1.2            0.96            -1.6             0.5            -1.1         -0.0055           -0.18           0.071            -0.3      5.2e+03    1.6e-06    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 42/100
Considering neighbor 0/20 for current solution
Attempt 43/100
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000268
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.71            0.23              -1               0               0               0               0               0           -0.31           -0.14               0               0      5.6e+03      0.044       10        1   ++
    1           -0.77             2.1            -1.6               0               0               0               0               0            0.24              -1               0               0      5.3e+03      0.023    1e+02     0.93   ++
    2           -0.88             1.9            -1.6               0               0               0               0               0            0.23            -1.3               0               0      5.2e+03    0.00045    1e+03        1   ++
    3           -0.88             1.9            -1.6               0               0               0               0               0            0.23            -1.3               0               0      5.2e+03    1.6e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000269
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.72            0.65           -0.95               2               0               0               0         -0.0015               0               0           -0.41            -0.3               0               0        6e+03        2.3        1     0.54    +
    1           -0.93             1.6           -0.38             1.9               0               0               0         -0.0073               0               0           -0.15           -0.73               0               0      5.5e+03       0.17        1     0.86    +
    2            -1.2             1.8           -0.83            0.89               0               0               0         -0.0047               0               0           -0.24           -0.85               0               0      5.3e+03      0.026       10        1   ++
    3           -0.51             1.8            -1.7           -0.05               0               0               0         -0.0059               0               0            0.19            -1.3               0               0      5.2e+03       0.12       10     0.49    +
    4           -0.51             1.9            -1.7            0.32               0               0               0          -0.006               0               0            0.18            -1.4               0               0      5.2e+03      0.015    1e+02        1   ++
    5            -0.6             1.9            -1.6            0.34               0               0               0          -0.006               0               0            0.12            -1.3               0               0      5.2e+03    0.00078    1e+03     0.98   ++
    6            -0.6             1.9            -1.6            0.34               0               0               0          -0.006               0               0            0.12            -1.3               0               0      5.2e+03    8.4e-07    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000270
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0            -0.9               1           -0.68               0               0               0               0               0           -0.38           -0.45               0               0      5.5e+03      0.041       10      1.1   ++
    1              -1             1.7            -1.1               0               0               0               0               0           0.053            -1.2               0               0      5.3e+03     0.0097    1e+02      1.1   ++
    2            -1.1             1.8            -1.1               0               0               0               0               0           0.057            -1.4               0               0      5.3e+03    0.00044    1e+03        1   ++
    3            -1.1             1.8            -1.1               0               0               0               0               0           0.057            -1.4               0               0      5.3e+03    1.2e-06    1e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000271
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.66             0.5              -1             1.7           -0.73               1         -0.0022           -0.46           -0.25      5.5e+03          2        1     0.76    +
    1            -0.9             1.5            -1.1             1.1              -1             0.9         -0.0058           -0.52           -0.65      5.1e+03       0.38       10      1.1   ++
    2           -0.45             2.1            -2.3          -0.071            -1.7           -0.25         -0.0061           -0.02            -1.7        5e+03      0.034       10     0.45    +
    3           -0.92             2.2            -1.6             0.1            -1.5           -0.11          -0.006           -0.28            -1.8      4.9e+03       0.03    1e+02        1   ++
    4           -0.88             2.2            -1.6            0.29            -1.5           0.035         -0.0062           -0.27            -1.9      4.9e+03     0.0066    1e+03     0.97   ++
    5            -0.9             2.2            -1.6            0.28            -1.5           0.043         -0.0062           -0.28            -1.8      4.9e+03      9e-05    1e+04        1   ++
    6            -0.9             2.2            -1.6            0.28            -1.5           0.043         -0.0062           -0.28            -1.8      4.9e+03    3.2e-06    1e+04        1   ++
Considering neighbor 3/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:boxcox [9859.55511216498, np.float64(9920.934760276556), 9]
Attempt 44/100
Considering neighbor 0/20 for current solution
Attempt 45/100
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000272
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0        6e+03        2.3        1     0.54    +
    1      5.4e+03       0.19       10        1   ++
    2      5.4e+03       0.19        5   -3e+03    -
    3      5.4e+03       0.19      2.5      -24    -
    4      5.4e+03       0.19      1.2    -0.64    -
    5      5.2e+03        0.1      1.2     0.83    +
    6      5.2e+03      0.079       12     0.96   ++
    7      5.2e+03     0.0015  1.2e+02        1   ++
    8      5.2e+03    1.6e-07  1.2e+02        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000273
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.36           -0.61           -0.83               2               0               0               0               0               0            -0.4           -0.18               0               0        6e+03       0.13        1     0.55    +
    1          -0.099            -1.6           -0.16               2               0               0               0               0               0           -0.64            0.41               0               0      5.7e+03      0.061        1     0.72    +
    2           -0.42            -1.3            -0.4            0.97               0               0               0               0               0           -0.67             0.2               0               0      5.6e+03      0.024       10     0.99   ++
    3            0.22            -1.2            -1.3            0.08               0               0               0               0               0            -0.2            0.31               0               0      5.4e+03      0.018    1e+02     0.99   ++
    4            0.49            -1.2            -1.7            0.63               0               0               0               0               0          -0.057            0.34               0               0      5.4e+03      0.019    1e+02     0.25    +
    5            0.43            -1.2            -1.6            0.43               0               0               0               0               0           -0.09            0.33               0               0      5.4e+03     0.0028    1e+03      1.1   ++
    6            0.42            -1.2            -1.6            0.39               0               0               0               0               0          -0.096            0.32               0               0      5.4e+03    0.00012    1e+04        1   ++
    7            0.42            -1.2            -1.6            0.39               0               0               0               0               0          -0.096            0.32               0               0      5.4e+03    2.7e-07    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000274
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.5e+03        2.9       10      1.1   ++
    1      5.5e+03        2.9        2      -98    -
    2      5.5e+03        2.9        1    0.016    -
    3      5.2e+03       0.54       10        1   ++
    4      5.2e+03       0.54        1     -6.2    -
    5      5.1e+03       0.22       10     0.92   ++
    6      5.1e+03      0.034    1e+02        1   ++
    7      5.1e+03    0.00087    1e+03        1   ++
    8      5.1e+03    2.2e-06    1e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000275
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03      0.037       10        1   ++
    1      5.7e+03     0.0061    1e+02      1.1   ++
    2      5.7e+03    0.00019    1e+03        1   ++
    3      5.7e+03    1.9e-07    1e+03        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000276
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03      0.037       10        1   ++
    1      5.7e+03     0.0061    1e+02      1.1   ++
    2      5.7e+03    0.00019    1e+03        1   ++
    3      5.7e+03    1.9e-07    1e+03        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000277
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.55           -0.94           -0.82               2               0               0               0          0.0077               0               0           -0.18            0.11               0               0        6e+03        2.7        1     0.65    +
    1           -0.55           -0.94           -0.82               2               0               0               0          0.0077               0               0           -0.18            0.11               0               0        6e+03        2.7      0.5    -0.64    -
    2          -0.047            -0.9           -0.46             1.8               0               0               0          -0.006               0               0           -0.43            0.11               0               0      5.6e+03       0.32        5        1   ++
    3          -0.047            -0.9           -0.46             1.8               0               0               0          -0.006               0               0           -0.43            0.11               0               0      5.6e+03       0.32      2.5      -22    -
    4          -0.047            -0.9           -0.46             1.8               0               0               0          -0.006               0               0           -0.43            0.11               0               0      5.6e+03       0.32      1.2    -0.83    -
    5          -0.078            -1.4            -1.2            0.54               0               0               0           0.004               0               0           -0.27            0.54               0               0      5.5e+03       0.41      1.2     0.78    +
    6            0.72            -1.2            -1.5            0.34               0               0               0         -0.0061               0               0           -0.23            0.32               0               0      5.4e+03       0.19       12     0.95   ++
    7            0.69            -1.3            -1.6            0.39               0               0               0         -0.0057               0               0           -0.21            0.31               0               0      5.4e+03     0.0054  1.2e+02        1   ++
    8            0.69            -1.3            -1.6            0.39               0               0               0         -0.0057               0               0           -0.21            0.31               0               0      5.4e+03    3.1e-06  1.2e+02        1   ++
Considering neighbor 5/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 19 unknown parameters [max: 50]
*** Estimate b07everything_000278
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03        2.9       10        1   ++
    1      5.7e+03       0.32    1e+02      1.1   ++
    2      5.7e+03      0.014    1e+03        1   ++
    3      5.7e+03    2.1e-05    1e+04        1   ++
    4      5.7e+03    4.6e-11    1e+04        1   ++
Considering neighbor 6/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 21 unknown parameters [max: 50]
*** Estimate b07everything_000279
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.6e+03        3.3       10      1.1   ++
    1      5.4e+03       0.55    1e+02      1.1   ++
    2      5.4e+03      0.048    1e+03        1   ++
    3      5.4e+03    0.00034    1e+04        1   ++
    4      5.4e+03    1.7e-08    1e+04        1   ++
Considering neighbor 7/20 for current solution
Considering neighbor 8/20 for current solution
Attempt 46/100
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000280
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time       lambda_tt          b_cost       b_headway         asc_car     Function    Relgrad   Radius      Rho
    0           -0.68           -0.99             1.8              -1         -0.0017           -0.46      5.7e+03        1.6        1     0.71    +
    1            -0.7            -1.7            0.83            -1.3         0.00048           -0.15      5.4e+03       0.19       10     0.92   ++
    2           -0.15            -1.7            0.59            -1.1         -0.0062           -0.13      5.3e+03      0.076    1e+02     0.98   ++
    3           -0.23            -1.7            0.51            -1.1         -0.0054           -0.11      5.3e+03     0.0013    1e+03        1   ++
    4           -0.23            -1.7            0.51            -1.1         -0.0054           -0.11      5.3e+03    6.9e-07    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000281
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost       b_headway     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.68            0.55              -1             1.8           -0.74         -0.0021           -0.46           -0.27      5.6e+03          2        1     0.69    +
    1           -0.93             1.6           -0.93             1.3            -1.2         -0.0064           -0.47           -0.74      5.1e+03       0.24       10      1.1   ++
    2           -0.93             1.6           -0.93             1.3            -1.2         -0.0064           -0.47           -0.74      5.1e+03       0.24      4.9 -2.4e+03    -
    3           -0.93             1.6           -0.93             1.3            -1.2         -0.0064           -0.47           -0.74      5.1e+03       0.24      2.5      -34    -
    4           -0.93             1.6           -0.93             1.3            -1.2         -0.0064           -0.47           -0.74      5.1e+03       0.24      1.2     -2.9    -
    5            -1.1             2.5            -1.9            0.12            -1.6         -0.0027           -0.15            -1.4        5e+03      0.081      1.2     0.78    +
    6           -0.87             2.2            -1.6            0.24            -1.5         -0.0064           -0.28            -1.8      4.9e+03     0.0087       12     0.98   ++
    7            -0.9             2.2            -1.6            0.27            -1.5         -0.0061           -0.28            -1.9      4.9e+03     0.0002  1.2e+02        1   ++
    8            -0.9             2.2            -1.6            0.27            -1.5         -0.0061           -0.28            -1.9      4.9e+03    5.2e-07  1.2e+02        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:log;train_headway_catalog:with_headway;train_tt_catalog:boxcox [9857.857591110185, np.float64(9912.417278320474), 8]
Attempt 47/100
Considering neighbor 0/20 for current solution
Attempt 48/100
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000282
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91               0               0               0            -0.3         -0.0074               0               0              -1               0               0      5.9e+03        2.6       10        1   ++
    1            -1.4               0               0               0              -2         -0.0057               0               0           -0.78               0               0      5.6e+03       0.22    1e+02     0.99   ++
    2            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.87               0               0      5.6e+03     0.0089    1e+03        1   ++
    3            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.88               0               0      5.6e+03    1.2e-05    1e+04        1   ++
    4            -1.4               0               0               0            -2.2         -0.0055               0               0           -0.88               0               0      5.6e+03    4.4e-11    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000283
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL         asc_car beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91               0               0               0           -0.22         -0.0073               0               0              -1               0               0      5.9e+03        2.6       10        1   ++
    1            -1.4               0               0               0            -1.1         -0.0057               0               0           -0.93               0               0      5.6e+03      0.042    1e+02     0.97   ++
    2            -1.4               0               0               0            -1.1         -0.0055               0               0           -0.98               0               0      5.6e+03    0.00046    1e+03        1   ++
    3            -1.4               0               0               0            -1.1         -0.0055               0               0           -0.98               0               0      5.6e+03    4.8e-08    1e+03        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 49/100
Considering neighbor 0/20 for current solution
Attempt 50/100
Considering neighbor 0/20 for current solution
Attempt 51/100
Considering neighbor 0/20 for current solution
Attempt 52/100
Considering neighbor 0/20 for current solution
Attempt 53/100
Considering neighbor 0/20 for current solution
Attempt 54/100
Considering neighbor 0/20 for current solution
Attempt 55/100
Considering neighbor 0/20 for current solution
Attempt 56/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000284
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.36           -0.57              -1           -0.74               1           -0.35           -0.15      5.4e+03      0.037       10        1   ++
    1            0.28            -1.3            -1.5            -1.3          -0.089           -0.39            0.24      5.3e+03      0.016       10      0.8    +
    2            0.27            -1.3            -1.6            -1.2            0.22           -0.41            0.29      5.2e+03     0.0039    1e+02      1.2   ++
    3             0.3            -1.3            -1.6            -1.1             0.4           -0.39            0.28      5.2e+03    0.00075    1e+03      1.1   ++
    4             0.3            -1.3            -1.6            -1.1            0.43           -0.38            0.28      5.2e+03    1.7e-05    1e+04        1   ++
    5             0.3            -1.3            -1.6            -1.1            0.43           -0.38            0.28      5.2e+03    2.4e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 57/100
Considering neighbor 0/20 for current solution
Attempt 58/100
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000285
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.5e+03        2.8       10        1   ++
    1      5.4e+03       0.53    1e+02      1.1   ++
    2      5.3e+03      0.044    1e+03      1.1   ++
    3      5.3e+03    0.00041    1e+04        1   ++
    4      5.3e+03    4.5e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000286
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC         asc_car beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.52              -1               0               0               0         -0.0026               0               0           -0.34               0               0      5.7e+03        2.5       10        1   ++
    1           -0.18            -1.5               0               0               0         -0.0048               0               0             0.1               0               0      5.6e+03       0.13    1e+02        1   ++
    2           -0.17            -1.6               0               0               0         -0.0053               0               0           0.091               0               0      5.6e+03     0.0022    1e+03        1   ++
    3           -0.17            -1.6               0               0               0         -0.0053               0               0           0.091               0               0      5.6e+03      5e-07    1e+03        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 59/100
Considering neighbor 0/20 for current solution
Attempt 60/100
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000287
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.43           -0.58            0.47              -1           -0.62               1           -0.37           -0.19           -0.21      5.3e+03       0.05       10        1   ++
    1           -0.43           -0.58            0.47              -1           -0.62               1           -0.37           -0.19           -0.21      5.3e+03       0.05      1.3       -3    -
    2           -0.32            -1.1             1.8            -1.6            -1.1            0.81           -0.38            0.15           -0.66      4.9e+03      0.019       13      1.1   ++
    3           -0.32            -1.1             1.8            -1.6            -1.1            0.81           -0.38            0.15           -0.66      4.9e+03      0.019     0.82     -2.4    -
    4           -0.26            -1.2             2.2            -1.7            -1.3          -0.011           -0.37            0.38            -1.1      4.8e+03     0.0087      8.2     0.99   ++
    5           -0.38            -1.2             2.1            -1.6            -1.5          -0.026           -0.55            0.46            -1.8      4.8e+03     0.0014       82      1.1   ++
    6           -0.38            -1.2             2.1            -1.6            -1.5           -0.04           -0.56            0.46              -2      4.8e+03    8.5e-05  8.2e+02        1   ++
    7           -0.38            -1.2             2.1            -1.6            -1.5           -0.04           -0.56            0.46              -2      4.8e+03    2.8e-07  8.2e+02        1   ++
Considering neighbor 0/20 for current solution
*** New pareto solution:
asc:MALE-GA;train_cost_catalog:boxcox;train_headway_catalog:without_headway;train_tt_catalog:log [9678.269631854551, np.float64(9739.649279966126), 9]
Attempt 61/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000288
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.52           0.052         -0.0022              -1               0               0               0         -0.0028               0               0           -0.25           -0.14          -0.034               0               0      5.6e+03        2.5       10        1   ++
    1            -0.7            0.81            0.63            -1.5               0               0               0         -0.0047               0               0            0.15          -0.089           -0.71               0               0      5.5e+03       0.44    1e+02      1.1   ++
    2           -0.94             1.1             0.9            -1.6               0               0               0         -0.0053               0               0            0.13          -0.081           -0.76               0               0      5.5e+03       0.05    1e+03      1.1   ++
    3           -0.98             1.1            0.94            -1.6               0               0               0         -0.0053               0               0            0.13          -0.082           -0.76               0               0      5.5e+03    0.00074    1e+04        1   ++
    4           -0.98             1.1            0.94            -1.6               0               0               0         -0.0053               0               0            0.13          -0.082           -0.76               0               0      5.5e+03      2e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000289
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.22            -0.6              -1               0               0               0         -0.0001               0               0           -0.25          0.0035               0               0      5.5e+03        2.7       10        1   ++
    1            0.67            -1.2            -1.5               0               0               0         -0.0046               0               0           -0.15            0.27               0               0      5.4e+03       0.46    1e+02        1   ++
    2            0.72            -1.3            -1.6               0               0               0         -0.0056               0               0           -0.17            0.27               0               0      5.4e+03      0.021    1e+03        1   ++
    3            0.72            -1.3            -1.6               0               0               0         -0.0057               0               0           -0.17            0.27               0               0      5.4e+03    4.4e-05    1e+04        1   ++
    4            0.72            -1.3            -1.6               0               0               0         -0.0057               0               0           -0.17            0.27               0               0      5.4e+03      2e-10    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000290
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0              -1           0.066          -0.017              -1               0               0               0               0               0           -0.39           -0.27          -0.046               0               0      5.7e+03      0.041       10        1   ++
    1           -0.97             0.8            0.62            -2.8               0               0               0               0               0            0.19          -0.051           -0.68               0               0      5.5e+03      0.014    1e+02      1.1   ++
    2            -1.2             1.1            0.85              -3               0               0               0               0               0            0.22          -0.075           -0.76               0               0      5.5e+03    0.00099    1e+03      1.1   ++
    3            -1.2             1.1            0.88              -3               0               0               0               0               0            0.22          -0.076           -0.77               0               0      5.5e+03    1.6e-05    1e+04        1   ++
    4            -1.2             1.1            0.88              -3               0               0               0               0               0            0.22          -0.076           -0.77               0               0      5.5e+03    4.3e-09    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000291
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03      0.055       10        1   ++
    1      5.2e+03      0.027    1e+02     0.99   ++
    2      5.2e+03     0.0017    1e+03      1.1   ++
    3      5.2e+03    4.1e-05    1e+04        1   ++
    4      5.2e+03    2.5e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000292
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.24           -0.56              -1               0               0               0               0               0           -0.26          -0.026               0               0      5.6e+03       0.04       10        1   ++
    1            0.45            -1.2            -1.5               0               0               0               0               0          -0.062            0.28               0               0      5.4e+03     0.0081    1e+02        1   ++
    2            0.45            -1.3            -1.6               0               0               0               0               0          -0.062            0.27               0               0      5.4e+03    0.00022    1e+03        1   ++
    3            0.45            -1.3            -1.6               0               0               0               0               0          -0.062            0.27               0               0      5.4e+03    2.4e-07    1e+03        1   ++
Considering neighbor 4/20 for current solution
Considering neighbor 5/20 for current solution
Attempt 62/100
Biogeme parameters read from biogeme.toml.
Model with 5 unknown parameters [max: 50]
*** Estimate b07everything_000293
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost       b_headway         asc_car     Function    Relgrad   Radius      Rho
    0           -0.72           -0.66           -0.95         -0.0037           -0.61      5.5e+03        2.2       10      1.1   ++
    1           -0.62              -1            -1.1         -0.0051           -0.41      5.4e+03        0.2    1e+02      1.1   ++
    2            -0.6            -1.1              -1         -0.0053           -0.38      5.4e+03     0.0056    1e+03        1   ++
    3            -0.6            -1.1              -1         -0.0054           -0.38      5.4e+03    0.00051    1e+04        1   ++
    4            -0.6            -1.1              -1         -0.0054           -0.38      5.4e+03      1e-05    1e+05        1   ++
    5            -0.6            -1.1              -1         -0.0054           -0.38      5.4e+03      4e-07    1e+05        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000294
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time       lambda_tt          b_cost       b_headway         asc_car     Function    Relgrad   Radius      Rho
    0           -0.72              -1               2           -0.85        -0.00082           -0.46      5.9e+03        1.6        1     0.56    +
    1              -1               0             2.1            -1.8         -0.0076           -0.84      5.8e+03       0.42        1     0.24    +
    2              -1               0             2.1            -1.8         -0.0076           -0.84      5.8e+03       0.42      0.5    -0.05    -
    3            -1.4          -0.039             2.3            -1.3         -0.0067            -1.1      5.6e+03       0.13        5     0.95   ++
    4            -1.4          -0.039             2.3            -1.3         -0.0067            -1.1      5.6e+03       0.13      2.5     -3.9    -
    5            -1.4          -0.039             2.3            -1.3         -0.0067            -1.1      5.6e+03       0.13      1.2    -0.59    -
    6            -1.4           -0.26               1            -0.7         -0.0003           -0.44      5.6e+03       0.31      1.2     0.38    +
    7           -0.76           -0.82               1           -0.99         -0.0051           -0.46      5.4e+03      0.041       12      1.1   ++
    8           -0.25            -1.7           0.064              -1         -0.0054           -0.14      5.4e+03      0.068       12     0.53    +
    9            -0.3            -1.6            0.37              -1         -0.0053           -0.18      5.4e+03     0.0066  1.2e+02        1   ++
   10           -0.37            -1.5             0.4              -1         -0.0053           -0.22      5.4e+03    0.00033  1.2e+03        1   ++
   11           -0.37            -1.5             0.4              -1         -0.0053           -0.22      5.4e+03    1.5e-06  1.2e+03        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 63/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000295
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.83               1           -0.71               0               0               0        -0.00098               0               0           -0.26           -0.43               0               0      5.4e+03        2.9       10      1.1   ++
    1           -0.83             1.7            -1.1               0               0               0         -0.0046               0               0          -0.054            -1.1               0               0      5.3e+03       0.58    1e+02      1.1   ++
    2           -0.85             1.8            -1.1               0               0               0          -0.006               0               0          -0.065            -1.3               0               0      5.3e+03      0.035    1e+03        1   ++
    3           -0.86             1.8            -1.1               0               0               0         -0.0061               0               0          -0.066            -1.3               0               0      5.3e+03    0.00014    1e+04        1   ++
    4           -0.86             1.8            -1.1               0               0               0         -0.0061               0               0          -0.066            -1.3               0               0      5.3e+03    2.2e-09    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 64/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000296
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      6.1e+03       0.07       10      1.1   ++
    1      5.9e+03     0.0023    1e+02        1   ++
    2      5.9e+03    1.1e-05    1e+03        1   ++
    3      5.9e+03    2.6e-10    1e+03        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 65/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000297
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC         asc_car beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.78           -0.65               0               0               0               0               0           -0.35               0               0      5.7e+03      0.035       10      1.1   ++
    1           -0.64            -1.1               0               0               0               0               0           0.032               0               0      5.6e+03     0.0056    1e+02        1   ++
    2           -0.63            -1.1               0               0               0               0               0           0.044               0               0      5.6e+03    8.2e-05    1e+03        1   ++
    3           -0.63            -1.1               0               0               0               0               0           0.044               0               0      5.6e+03    1.9e-08    1e+03        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 66/100
Considering neighbor 0/20 for current solution
Attempt 67/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000298
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.52           0.052         -0.0022              -1               0               0               0         -0.0028               0               0           -0.25           -0.14          -0.034               0               0      5.6e+03        2.5       10        1   ++
    1            -0.7            0.81            0.63            -1.5               0               0               0         -0.0047               0               0            0.15          -0.089           -0.71               0               0      5.5e+03       0.44    1e+02      1.1   ++
    2           -0.94             1.1             0.9            -1.6               0               0               0         -0.0053               0               0            0.13          -0.081           -0.76               0               0      5.5e+03       0.05    1e+03      1.1   ++
    3           -0.98             1.1            0.94            -1.6               0               0               0         -0.0053               0               0            0.13          -0.082           -0.76               0               0      5.5e+03    0.00074    1e+04        1   ++
    4           -0.98             1.1            0.94            -1.6               0               0               0         -0.0053               0               0            0.13          -0.082           -0.76               0               0      5.5e+03      2e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000299
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042       10      1.1   ++
    1              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042        5 -1.4e+07    -
    2              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042      2.5 -2.1e+02    -
    3              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042      1.2    -0.26    -
    4            -1.4             1.4            -1.9            -1.1            0.99           -0.38            -1.1      5.1e+03      0.023       12      1.1   ++
    5            -1.4             1.4            -1.9            -1.1            0.99           -0.38            -1.1      5.1e+03      0.023      1.2      -12    -
    6            -1.3             2.5              -3              -2           -0.19           -0.31            -1.4        5e+03      0.041      1.2      0.5    +
    7            -1.3             2.1            -2.9            -1.4           -0.13           -0.18            -1.8        5e+03     0.0015       12     0.98   ++
    8            -1.2             2.1              -3            -1.5             0.1           -0.19            -1.9      4.9e+03      0.002  1.2e+02     0.91   ++
    9            -1.2             2.1              -3            -1.5           0.091           -0.19            -1.9      4.9e+03    1.9e-05  1.2e+03        1   ++
   10            -1.2             2.1              -3            -1.5           0.091           -0.19            -1.9      4.9e+03    2.3e-08  1.2e+03        1   ++
Considering neighbor 1/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:boxcox;train_headway_catalog:without_headway;train_tt_catalog:sqrt [9906.475472994323, np.float64(9954.215199303326), 7]
Attempt 68/100
Considering neighbor 0/20 for current solution
Attempt 69/100
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000300
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.68           -0.11          -0.016            0.24              -1           -0.29               1           -0.34           -0.26          -0.021           -0.14      5.5e+03      0.055       10        1   ++
    1           -0.68           -0.11          -0.016            0.24              -1           -0.29               1           -0.34           -0.26          -0.021           -0.14      5.5e+03      0.055      4.5 -2.2e+05    -
    2           -0.68           -0.11          -0.016            0.24              -1           -0.29               1           -0.34           -0.26          -0.021           -0.14      5.5e+03      0.055      2.2      -88    -
    3           -0.68           -0.11          -0.016            0.24              -1           -0.29               1           -0.34           -0.26          -0.021           -0.14      5.5e+03      0.055      1.1     -2.8    -
    4              -1            0.49          0.0064             1.4            -1.6            -1.2               1          -0.036           -0.14          -0.077           -0.49      5.1e+03      0.018       11        1   ++
    5              -1            0.49          0.0064             1.4            -1.6            -1.2               1          -0.036           -0.14          -0.077           -0.49      5.1e+03      0.018     0.99     -4.4    -
    6            -1.5            0.62            0.15             2.4            -1.7            -1.7            0.25          -0.099          -0.032           -0.22            -1.1      4.9e+03      0.013      9.9        1   ++
    7            -1.7            0.71            0.47               2            -1.6            -1.5           0.016           -0.19           0.046           -0.41            -1.8      4.9e+03     0.0019       99        1   ++
    8            -1.7            0.71            0.45               2            -1.6            -1.5          -0.035           -0.18           0.044           -0.45            -1.8      4.9e+03    9.4e-05  9.9e+02        1   ++
    9            -1.7            0.71            0.45               2            -1.6            -1.5          -0.035           -0.18           0.044           -0.45            -1.8      4.9e+03    9.2e-08  9.9e+02        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 70/100
Considering neighbor 0/20 for current solution
Attempt 71/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000301
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.53           -0.81            0.78              -1               0               0               0         0.00034               0               0           -0.42          -0.079           -0.32               0               0      5.4e+03        3.2       10      1.1   ++
    1            0.04           -0.98             1.6            -2.8               0               0               0         -0.0045               0               0           -0.25            0.38            -1.2               0               0      5.1e+03       0.64    1e+02      1.1   ++
    2            0.16            -1.1             1.7              -3               0               0               0         -0.0063               0               0           -0.29            0.42            -1.5               0               0      5.1e+03      0.059    1e+03      1.1   ++
    3            0.17            -1.1             1.8              -3               0               0               0         -0.0065               0               0           -0.29            0.43            -1.5               0               0      5.1e+03    0.00056    1e+04        1   ++
    4            0.17            -1.1             1.8              -3               0               0               0         -0.0065               0               0           -0.29            0.43            -1.5               0               0      5.1e+03    4.8e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 72/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000302
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.53           -0.81            0.78              -1               0               0               0         0.00034               0               0           -0.42          -0.079           -0.32               0               0      5.4e+03        3.2       10      1.1   ++
    1            0.04           -0.98             1.6            -2.8               0               0               0         -0.0045               0               0           -0.25            0.38            -1.2               0               0      5.1e+03       0.64    1e+02      1.1   ++
    2            0.16            -1.1             1.7              -3               0               0               0         -0.0063               0               0           -0.29            0.42            -1.5               0               0      5.1e+03      0.059    1e+03      1.1   ++
    3            0.17            -1.1             1.8              -3               0               0               0         -0.0065               0               0           -0.29            0.43            -1.5               0               0      5.1e+03    0.00056    1e+04        1   ++
    4            0.17            -1.1             1.8              -3               0               0               0         -0.0065               0               0           -0.29            0.43            -1.5               0               0      5.1e+03    4.8e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 23 unknown parameters [max: 50]
*** Estimate b07everything_000303
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03          3       10        1   ++
    1      5.6e+03       0.41    1e+02      1.1   ++
    2      5.6e+03      0.034    1e+03      1.1   ++
    3      5.6e+03    0.00026    1e+04        1   ++
    4      5.6e+03    2.1e-08    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000304
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.59           -0.77            0.75              -1           -0.86           -0.46           -0.18           -0.28      5.3e+03      0.047       10      1.1   ++
    1           -0.43           -0.97             1.8            -2.7            -2.5            -0.5            0.37            0.88      4.9e+03      0.019    1e+02      1.1   ++
    2           -0.37            -1.1               2            -3.1            -2.7           -0.54            0.44               1      4.8e+03     0.0014    1e+03        1   ++
    3           -0.36            -1.1               2            -3.1            -2.8           -0.54            0.44               1      4.8e+03    7.5e-06    1e+04        1   ++
    4           -0.36            -1.1               2            -3.1            -2.8           -0.54            0.44               1      4.8e+03    2.5e-10    1e+04        1   ++
Considering neighbor 2/20 for current solution
*** New pareto solution:
asc:MALE-GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:sqrt [9708.681955643766, np.float64(9763.241642854055), 8]
Attempt 73/100
Biogeme parameters read from biogeme.toml.
Model with 21 unknown parameters [max: 50]
*** Estimate b07everything_000305
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03        2.5       10        1   ++
    1      5.8e+03       0.56    1e+02      1.1   ++
    2      5.7e+03       0.07    1e+03      1.1   ++
    3      5.7e+03     0.0016    1e+04        1   ++
    4      5.7e+03    9.4e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000306
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_ma asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.87           -0.95            0.75               0               0               0              -1               0               0           -0.69           -0.37           -0.35               0               0      5.4e+03      0.041       10        1   ++
    1            -1.4           -0.99             2.1               0               0               0            -1.4               0               0            -1.2             0.3            -1.3               0               0      5.1e+03      0.017    1e+02      1.1   ++
    2            -1.5            -1.2             2.2               0               0               0            -1.5               0               0            -1.3            0.34            -1.6               0               0      5.1e+03      0.001    1e+03        1   ++
    3            -1.5            -1.2             2.2               0               0               0            -1.5               0               0            -1.3            0.35            -1.7               0               0      5.1e+03    6.4e-06    1e+04        1   ++
    4            -1.5            -1.2             2.2               0               0               0            -1.5               0               0            -1.3            0.35            -1.7               0               0      5.1e+03    8.5e-10    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000307
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_ma asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.83              -1            0.49               0               0               0           -0.66         0.00091               0               0           -0.59           -0.29           -0.24               0               0      5.5e+03          3       10      1.1   ++
    1            -1.2           -0.99             2.2               0               0               0            -1.4         -0.0048               0               0            -1.3            0.29            -1.2               0               0      5.1e+03       0.89    1e+02        1   ++
    2            -1.2            -1.2             2.2               0               0               0            -1.5         -0.0067               0               0            -1.4            0.34            -1.6               0               0      5.1e+03      0.066    1e+03        1   ++
    3            -1.2            -1.2             2.2               0               0               0            -1.5         -0.0069               0               0            -1.4            0.35            -1.7               0               0      5.1e+03    0.00066    1e+04        1   ++
    4            -1.2            -1.2             2.2               0               0               0            -1.5         -0.0069               0               0            -1.4            0.35            -1.7               0               0      5.1e+03    6.1e-08    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000308
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.66          -0.081          -0.015              -1           -0.29           -0.35           -0.27          -0.023      5.6e+03      0.058       10        1   ++
    1            -1.1             0.8            0.64            -1.5            -1.1          -0.072          -0.072           -0.68      5.3e+03      0.013    1e+02        1   ++
    2            -1.4             1.1            0.94            -1.5            -1.1          -0.083          -0.081           -0.72      5.3e+03     0.0014    1e+03      1.1   ++
    3            -1.5             1.2            0.99            -1.5            -1.1          -0.082          -0.083           -0.73      5.3e+03    2.9e-05    1e+04        1   ++
    4            -1.5             1.2            0.99            -1.5            -1.1          -0.082          -0.083           -0.73      5.3e+03    1.1e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000309
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.43           -0.71            0.12         -0.0019              -1           -0.67           0.002           -0.21         -0.0063           -0.14          -0.037      5.3e+03        2.6       10      1.1   ++
    1           -0.12            -1.1            0.73            0.33            -1.5            -1.1         -0.0043           -0.41            0.24          -0.069           -0.51      5.2e+03       0.64    1e+02      1.1   ++
    2           -0.22            -1.2            0.94             0.5            -1.5            -1.1         -0.0057           -0.45            0.27          -0.079           -0.62      5.2e+03      0.055    1e+03      1.1   ++
    3           -0.24            -1.2            0.97            0.52            -1.5            -1.1         -0.0058           -0.45            0.27           -0.08           -0.63      5.2e+03    0.00059    1e+04        1   ++
    4           -0.24            -1.2            0.97            0.52            -1.5            -1.1         -0.0058           -0.45            0.27           -0.08           -0.63      5.2e+03    1.2e-07    1e+04        1   ++
Considering neighbor 4/20 for current solution
Considering neighbor 5/20 for current solution
Attempt 74/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000310
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0            -0.2           -0.82            -0.6           -0.93         -0.0028           -0.63          -0.038      5.4e+03        2.6       10      1.1   ++
    1            0.15            -1.2           -0.98              -1         -0.0051           -0.71            0.32      5.3e+03       0.36    1e+02      1.1   ++
    2            0.21            -1.2              -1            -1.1         -0.0058           -0.74            0.35      5.3e+03      0.015    1e+03        1   ++
    3            0.21            -1.2              -1            -1.1         -0.0058           -0.74            0.36      5.3e+03    2.9e-05    1e+04        1   ++
    4            0.21            -1.2              -1            -1.1         -0.0058           -0.74            0.36      5.3e+03    1.1e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000311
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.62           0.017          -0.015            0.36              -1             1.5           -0.41           -0.42           -0.32          -0.029           -0.19      5.7e+03      0.088        1     0.85    +
    1            -1.2            0.38          0.0051             1.3            -1.3             0.9            -1.4            0.16           0.092          -0.078           -0.39      5.1e+03      0.032       10     0.95   ++
    2            -1.4            0.67            0.49             1.9            -1.9            0.22            -2.6          -0.002           0.048           -0.35            0.95      4.9e+03     0.0098    1e+02     0.97   ++
    3            -1.6            0.71            0.49             1.9            -1.6            0.31            -2.8           -0.15           0.048           -0.36             1.2      4.9e+03      0.001    1e+03        1   ++
    4            -1.6            0.71            0.49             1.9            -1.6            0.33            -2.8           -0.15           0.049           -0.36             1.2      4.9e+03    2.6e-05    1e+04        1   ++
    5            -1.6            0.71            0.49             1.9            -1.6            0.33            -2.8           -0.15           0.049           -0.36             1.2      4.9e+03    2.7e-09    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000312
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.53          -0.046         -0.0064            0.33              -1            -0.4          -0.004           -0.38           -0.22          -0.029           -0.16      5.4e+03        2.5       10        1   ++
    1            -1.1            0.41            0.16             2.2            -1.5            -1.3         -0.0052           -0.27           0.056           -0.23            -1.3      4.9e+03          1    1e+02        1   ++
    2            -1.3            0.65            0.35               2            -1.6            -1.5          -0.006           -0.31           0.047           -0.42            -1.8      4.9e+03       0.07    1e+03        1   ++
    3            -1.4            0.71            0.42             2.1            -1.6            -1.5         -0.0061           -0.31           0.045           -0.45            -1.8      4.9e+03     0.0018    1e+04        1   ++
    4            -1.4            0.71            0.42             2.1            -1.6            -1.5         -0.0061           -0.31           0.045           -0.45            -1.8      4.9e+03    1.2e-06    1e+04        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 75/100
Biogeme parameters read from biogeme.toml.
Model with 8 unknown parameters [max: 50]
*** Estimate b07everything_000313
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.44           -0.59            0.47              -1           -0.76           -0.34           -0.17           -0.18      5.2e+03      0.044       10        1   ++
    1           -0.21              -1               2            -1.6              -1           -0.36            0.33           -0.35      4.9e+03      0.018    1e+02        1   ++
    2            -0.2            -1.2               2            -1.7            -1.1           -0.39            0.37           -0.42      4.9e+03     0.0011    1e+03        1   ++
    3            -0.2            -1.2               2            -1.7            -1.1           -0.39            0.38           -0.42      4.9e+03    8.5e-06    1e+04        1   ++
    4            -0.2            -1.2               2            -1.7            -1.1           -0.39            0.38           -0.42      4.9e+03    5.5e-10    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 76/100
Considering neighbor 0/20 for current solution
Attempt 77/100
Biogeme parameters read from biogeme.toml.
Model with 20 unknown parameters [max: 50]
*** Estimate b07everything_000314
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0        6e+03      0.041       10        1   ++
    1      5.8e+03      0.023    1e+02      1.1   ++
    2      5.8e+03     0.0038    1e+03      1.1   ++
    3      5.8e+03    0.00012    1e+04        1   ++
    4      5.8e+03    1.3e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 20 unknown parameters [max: 50]
*** Estimate b07everything_000315
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0        6e+03      0.041       10        1   ++
    1      5.8e+03      0.023    1e+02      1.1   ++
    2      5.8e+03     0.0038    1e+03      1.1   ++
    3      5.8e+03    0.00012    1e+04        1   ++
    4      5.8e+03    1.3e-07    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000316
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0        6e+03      0.073       10      1.1   ++
    1      5.5e+03      0.031    1e+02     0.98   ++
    2      5.5e+03     0.0011    1e+03        1   ++
    3      5.5e+03    4.7e-06    1e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 9 unknown parameters [max: 50]
*** Estimate b07everything_000317
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.92          -0.039          -0.011              -1           -0.64         -0.0019           -0.46            -0.2          -0.047      5.6e+03        2.5       10        1   ++
    1           -0.89            0.82             0.5            -2.9              -2         -0.0046           -0.16            0.17           -0.14      5.2e+03       0.39    1e+02      1.1   ++
    2            -1.1             1.1             0.9            -3.2            -2.3         -0.0055           -0.17            0.17           -0.24      5.1e+03       0.02    1e+03      1.1   ++
    3            -1.1             1.2            0.95            -3.2            -2.4         -0.0055           -0.18            0.17           -0.23      5.1e+03    0.00054    1e+04        1   ++
    4            -1.1             1.2            0.95            -3.2            -2.4         -0.0055           -0.18            0.17           -0.22      5.1e+03     0.0001    1e+05        1   ++
    5            -1.1             1.2            0.95            -3.2            -2.4         -0.0055           -0.18            0.17           -0.22      5.1e+03      6e-09    1e+05        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 19 unknown parameters [max: 50]
*** Estimate b07everything_000318
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03        2.7       10        1   ++
    1      5.5e+03       0.93    1e+02     0.97   ++
    2      5.5e+03      0.053    1e+03        1   ++
    3      5.5e+03    0.00034    1e+04        1   ++
    4      5.5e+03    1.4e-08    1e+04        1   ++
Considering neighbor 4/20 for current solution
Considering neighbor 5/20 for current solution
Attempt 78/100
Considering neighbor 0/20 for current solution
Attempt 79/100
Considering neighbor 0/20 for current solution
Attempt 80/100
Considering neighbor 0/20 for current solution
Attempt 81/100
Considering neighbor 0/20 for current solution
Attempt 82/100
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000319
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.4e+03          3       10      1.1   ++
    1      5.2e+03       0.64    1e+02      1.1   ++
    2      5.2e+03       0.05    1e+03      1.1   ++
    3      5.2e+03    0.00049    1e+04        1   ++
    4      5.2e+03    6.3e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000320
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03      0.087        1     0.74    +
    1      5.3e+03      0.024       10     0.98   ++
    2      5.3e+03      0.043       10     0.36    +
    3      5.2e+03     0.0014    1e+02      1.1   ++
    4      5.2e+03     0.0017    1e+03     0.99   ++
    5      5.2e+03    9.9e-06    1e+04        1   ++
    6      5.2e+03    3.1e-10    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000321
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.59           -0.97            0.73           -0.88               2           -0.87          0.0066            -0.2           0.088            -0.3      5.6e+03        2.9        1      0.7    +
    1           -0.29            -1.1             1.7           -0.49             1.8            -1.9          -0.008           -0.61            0.12           -0.39      5.1e+03       0.61       10        1   ++
    2           -0.29            -1.1             1.7           -0.49             1.8            -1.9          -0.008           -0.61            0.12           -0.39      5.1e+03       0.61        5 -1.7e+03    -
    3           -0.29            -1.1             1.7           -0.49             1.8            -1.9          -0.008           -0.61            0.12           -0.39      5.1e+03       0.61      2.5      -14    -
    4           -0.29            -1.1             1.7           -0.49             1.8            -1.9          -0.008           -0.61            0.12           -0.39      5.1e+03       0.61      1.2    -0.42    -
    5           -0.59            -1.3             1.8            -1.2            0.54            -2.2          0.0012           -0.52            0.48           -0.23      4.9e+03       0.19      1.2     0.87    +
    6            0.11            -1.2             1.9            -1.7             0.2            -2.7         -0.0069           -0.55            0.38               1      4.8e+03      0.035       12     0.93   ++
    7           0.028            -1.2               2            -1.7            0.29            -2.8         -0.0067           -0.63            0.43               1      4.8e+03     0.0025  1.2e+02        1   ++
    8           0.028            -1.2               2            -1.7            0.29            -2.8         -0.0067           -0.63            0.43               1      4.8e+03    3.9e-06  1.2e+02        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000322
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.74           0.083         -0.0053            0.69              -1               2           -0.78         -0.0018            -0.3           -0.15          -0.049           -0.28      5.8e+03        2.2        1     0.58    +
    1            -1.2            0.49           0.044             1.6           -0.68             1.7            -1.8          -0.008           -0.31           0.042           -0.11           -0.41      5.2e+03        0.1       10        1   ++
    2            -1.2            0.49           0.044             1.6           -0.68             1.7            -1.8          -0.008           -0.31           0.042           -0.11           -0.41      5.2e+03        0.1        5 -3.3e+03    -
    3            -1.2            0.49           0.044             1.6           -0.68             1.7            -1.8          -0.008           -0.31           0.042           -0.11           -0.41      5.2e+03        0.1      2.5      -29    -
    4            -1.2            0.49           0.044             1.6           -0.68             1.7            -1.8          -0.008           -0.31           0.042           -0.11           -0.41      5.2e+03        0.1      1.2    -0.75    -
    5            -1.5            0.54            0.12             1.8            -1.6            0.47            -2.2        -0.00014           -0.22            0.39           -0.15           -0.23        5e+03      0.059      1.2     0.88    +
    6            -1.1            0.66            0.53             1.9            -1.7            0.35            -2.7         -0.0071           -0.26           0.081           -0.32               1      4.9e+03      0.046       12     0.96   ++
    7            -1.3            0.71            0.47             1.9            -1.6            0.34            -2.8         -0.0062           -0.28           0.051           -0.36             1.2      4.9e+03     0.0038  1.2e+02        1   ++
    8            -1.3            0.71            0.47             1.9            -1.6            0.34            -2.8         -0.0062           -0.28           0.051           -0.36             1.2      4.9e+03    6.6e-06  1.2e+03        1   ++
    9            -1.3            0.71            0.47             1.9            -1.6            0.34            -2.8         -0.0062           -0.28           0.051           -0.36             1.2      4.9e+03    1.6e-08  1.2e+03        1   ++
Considering neighbor 3/20 for current solution
Considering neighbor 4/20 for current solution
Attempt 83/100
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000323
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.94          -0.088          -0.013            0.54              -1           -0.59         -0.0026            -0.5           -0.22          -0.044           -0.22      5.5e+03        2.6       10        1   ++
    1            -1.1            0.41            0.23             1.9            -2.6            -2.5         -0.0049           -0.31           0.083            -0.2               1      4.9e+03       0.75    1e+02        1   ++
    2            -1.3            0.66            0.45             1.9              -3            -2.8         -0.0061           -0.32           0.066           -0.34             1.2      4.9e+03      0.071    1e+03      1.1   ++
    3            -1.3            0.71            0.49             1.9            -3.1            -2.8         -0.0063           -0.32           0.064           -0.36             1.2      4.9e+03     0.0013    1e+04        1   ++
    4            -1.3            0.71            0.49             1.9            -3.1            -2.8         -0.0063           -0.32           0.064           -0.36             1.2      4.9e+03    4.5e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000324
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC         asc_car beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.65           -0.65               0               0               0         -0.0026               0               0            -0.4               0               0      5.7e+03        2.5       10      1.1   ++
    1           -0.41            -1.1               0               0               0         -0.0047               0               0          -0.064               0               0      5.6e+03       0.22    1e+02      1.1   ++
    2           -0.39            -1.1               0               0               0         -0.0052               0               0          -0.062               0               0      5.6e+03     0.0049    1e+03        1   ++
    3           -0.39            -1.1               0               0               0         -0.0052               0               0          -0.062               0               0      5.6e+03    2.4e-06    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 23 unknown parameters [max: 50]
*** Estimate b07everything_000325
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03          3       10        1   ++
    1      5.6e+03       0.41    1e+02      1.1   ++
    2      5.6e+03      0.034    1e+03      1.1   ++
    3      5.6e+03    0.00026    1e+04        1   ++
    4      5.6e+03    2.1e-08    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000326
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0            -0.8            0.19         -0.0025             0.9           -0.84              -1         -0.0018           -0.16          -0.055          -0.057           -0.33      5.2e+03        2.7       10      1.1   ++
    1            -1.4            0.52            0.38             1.7              -1            -2.5         -0.0049           -0.46           0.093           -0.22            0.99        5e+03       0.63    1e+02      1.1   ++
    2            -1.5            0.69            0.54             1.9            -1.1            -2.8         -0.0062            -0.5             0.1           -0.31             1.2        5e+03      0.054    1e+03      1.1   ++
    3            -1.5            0.72            0.56             1.9            -1.1            -2.8         -0.0063            -0.5             0.1           -0.31             1.2        5e+03    0.00056    1e+04        1   ++
    4            -1.5            0.72            0.56             1.9            -1.1            -2.8         -0.0063            -0.5             0.1           -0.31             1.2        5e+03    9.2e-08    1e+04        1   ++
Considering neighbor 3/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 23 unknown parameters [max: 50]
*** Estimate b07everything_000327
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03        2.6       10        1   ++
    1      5.5e+03       0.94    1e+02        1   ++
    2      5.5e+03      0.089    1e+03      1.1   ++
    3      5.5e+03     0.0022    1e+04        1   ++
    4      5.5e+03    1.6e-06    1e+04        1   ++
Considering neighbor 4/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000328
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_on asc_car_diff_se     Function    Relgrad   Radius      Rho
    0           -0.57           -0.89            0.24         -0.0075              -1           -0.88         0.00074           -0.36          -0.062           -0.15          -0.054      5.4e+03        2.9       10      1.1   ++
    1           -0.19            -1.1            0.75            0.35            -2.8            -2.1         -0.0043           -0.46            0.29            0.18           0.021        5e+03       0.43    1e+02      1.1   ++
    2           -0.18            -1.1            0.94            0.52            -3.2            -2.4         -0.0057           -0.53            0.36            0.19         -0.0019        5e+03      0.037    1e+03        1   ++
    3           -0.19            -1.1            0.96            0.55            -3.2            -2.4         -0.0059           -0.54            0.37            0.19          0.0044        5e+03    0.00036    1e+04        1   ++
    4           -0.19            -1.1            0.96            0.55            -3.2            -2.4         -0.0059           -0.54            0.37            0.19          0.0044        5e+03    3.5e-07    1e+04        1   ++
Considering neighbor 5/20 for current solution
Considering neighbor 6/20 for current solution
Attempt 84/100
Biogeme parameters read from biogeme.toml.
Model with 17 unknown parameters [max: 50]
*** Estimate b07everything_000329
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03        2.5       10        1   ++
    1      5.2e+03       0.86    1e+02        1   ++
    2      5.2e+03      0.085    1e+03      1.1   ++
    3      5.2e+03      0.002    1e+04        1   ++
    4      5.2e+03    1.4e-06    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 85/100
Considering neighbor 0/20 for current solution
Attempt 86/100
Considering neighbor 0/20 for current solution
Attempt 87/100
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000330
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.67            -0.1          -0.016            0.24              -1           -0.28           -0.34           -0.26          -0.021           -0.14      5.6e+03      0.065       10        1   ++
    1            -1.3            0.39             0.2             2.2            -1.5            -2.7           -0.14            0.07           -0.18             1.3        5e+03      0.031    1e+02     0.95   ++
    2            -1.5            0.66            0.42               2            -1.6            -2.8           -0.15           0.033           -0.34             1.2      4.9e+03     0.0017    1e+03        1   ++
    3            -1.6            0.71            0.47               2            -1.6            -2.8           -0.15           0.029           -0.35             1.2      4.9e+03    3.9e-05    1e+04        1   ++
    4            -1.6            0.71            0.47               2            -1.6            -2.8           -0.15           0.029           -0.35             1.2      4.9e+03    1.9e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 11 unknown parameters [max: 50]
*** Estimate b07everything_000331
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.94           -0.12          -0.015            0.51              -1           -0.79         -0.0032           -0.59           -0.27          -0.042           -0.21      5.4e+03        2.3       10        1   ++
    1              -1            0.41            0.22             1.9            -2.8              -1          -0.005           -0.27           0.066           -0.21           -0.36        5e+03       0.97    1e+02        1   ++
    2            -1.2            0.66            0.45             1.9            -3.2            -1.1         -0.0061           -0.24            0.04            -0.3           -0.27        5e+03      0.084    1e+03      1.1   ++
    3            -1.2            0.71             0.5             1.9            -3.2            -1.1         -0.0062           -0.24           0.039            -0.3           -0.27        5e+03     0.0016    1e+04        1   ++
    4            -1.2            0.71             0.5             1.9            -3.2            -1.1         -0.0062           -0.24           0.039            -0.3           -0.27        5e+03    7.4e-07    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000332
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.51            -0.8              -1           -0.93         -0.0005           -0.54           -0.19      5.4e+03        2.6       10      1.1   ++
    1            0.29            -1.1            -2.6              -1         -0.0046           -0.59            0.34      5.2e+03       0.36    1e+02      1.1   ++
    2            0.44            -1.3            -2.9            -1.1         -0.0057            -0.6            0.36      5.2e+03      0.021    1e+03        1   ++
    3            0.45            -1.3            -2.9            -1.1         -0.0058           -0.59            0.35      5.2e+03    5.6e-05    1e+04        1   ++
    4            0.45            -1.3            -2.9            -1.1         -0.0058           -0.59            0.35      5.2e+03    2.1e-07    1e+04        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000333
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03       0.04       10      1.1   ++
    1      5.2e+03      0.059    1e+02        1   ++
    2      5.2e+03     0.0037    1e+03        1   ++
    3      5.2e+03    0.00017    1e+04        1   ++
    4      5.2e+03      2e-07    1e+04        1   ++
Considering neighbor 3/20 for current solution
Considering neighbor 4/20 for current solution
Attempt 88/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000334
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0          -0.083           -0.77           -0.62               0               0               0         -0.0027               0               0           -0.47           0.023               0               0      5.6e+03        2.9       10      1.1   ++
    1            0.34            -1.1            -1.1               0               0               0          -0.005               0               0           -0.38            0.33               0               0      5.4e+03       0.37    1e+02      1.1   ++
    2            0.41            -1.2            -1.1               0               0               0         -0.0057               0               0           -0.39            0.32               0               0      5.4e+03      0.018    1e+03        1   ++
    3            0.41            -1.2            -1.1               0               0               0         -0.0057               0               0           -0.39            0.32               0               0      5.4e+03    3.6e-05    1e+04        1   ++
    4            0.41            -1.2            -1.1               0               0               0         -0.0057               0               0           -0.39            0.32               0               0      5.4e+03    1.4e-10    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 89/100
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000335
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0              -1           0.036              -1               0               0               0               0               0         -0.0065              -1               0               0      5.7e+03      0.038       10      1.1   ++
    1           -0.95             2.1            -2.6               0               0               0               0               0            0.14            -1.2               0               0      5.3e+03      0.027    1e+02     0.92   ++
    2           -0.91             1.8              -3               0               0               0               0               0            0.21            -1.4               0               0      5.2e+03    0.00075    1e+03        1   ++
    3           -0.91             1.8              -3               0               0               0               0               0            0.21            -1.4               0               0      5.2e+03    1.9e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000336
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.65           0.022           -0.01              -1               0               0               0               0               0           -0.22            -0.2          -0.029               0               0      5.7e+03      0.045       10        1   ++
    1           -0.91             0.8            0.63            -1.5               0               0               0               0               0            0.25          -0.088           -0.72               0               0      5.5e+03      0.009    1e+02      1.1   ++
    2            -1.2             1.1             0.9            -1.6               0               0               0               0               0            0.24          -0.083           -0.77               0               0      5.5e+03      0.001    1e+03      1.1   ++
    3            -1.2             1.1            0.94            -1.6               0               0               0               0               0            0.24          -0.084           -0.77               0               0      5.5e+03    1.8e-05    1e+04        1   ++
    4            -1.2             1.1            0.94            -1.6               0               0               0               0               0            0.24          -0.084           -0.77               0               0      5.5e+03      5e-09    1e+04        1   ++
Considering neighbor 1/20 for current solution
Considering neighbor 2/20 for current solution
Attempt 90/100
Biogeme parameters read from biogeme.toml.
Model with 7 unknown parameters [max: 50]
*** Estimate b07everything_000337
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost       b_headway     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.48           -0.77              -1           -0.99         -0.0016           -0.55           -0.21      5.4e+03        2.4       10        1   ++
    1            0.49            -1.2              -3              -1         -0.0048           -0.39            0.26      5.1e+03       0.35    1e+02      1.1   ++
    2            0.62            -1.3            -3.3            -1.1         -0.0057           -0.42            0.31      5.1e+03      0.017    1e+03        1   ++
    3            0.63            -1.3            -3.3            -1.1         -0.0058           -0.41             0.3      5.1e+03    6.3e-05    1e+04        1   ++
    4            0.63            -1.3            -3.3            -1.1         -0.0058           -0.41             0.3      5.1e+03    3.6e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000338
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_ma     Function    Relgrad   Radius      Rho
    0           -0.37           -0.59              -1           -0.89           -0.33           -0.13      5.3e+03      0.036       10        1   ++
    1             0.4            -1.3            -1.6              -1           -0.22            0.22      5.2e+03     0.0091    1e+02        1   ++
    2             0.4            -1.3            -1.7              -1           -0.24            0.25      5.2e+03    0.00022    1e+03        1   ++
    3             0.4            -1.3            -1.7              -1           -0.24            0.25      5.2e+03    2.4e-07    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000339
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_ma beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.24           -0.56              -1               0               0               0               0               0           -0.26          -0.026               0               0      5.6e+03       0.04       10        1   ++
    1            0.45            -1.2            -1.5               0               0               0               0               0          -0.062            0.28               0               0      5.4e+03     0.0081    1e+02        1   ++
    2            0.45            -1.3            -1.6               0               0               0               0               0          -0.062            0.27               0               0      5.4e+03    0.00022    1e+03        1   ++
    3            0.45            -1.3            -1.6               0               0               0               0               0          -0.062            0.27               0               0      5.4e+03    2.4e-07    1e+03        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 91/100
Biogeme parameters read from biogeme.toml.
Model with 15 unknown parameters [max: 50]
*** Estimate b07everything_000340
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_on asc_car_diff_se beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.67            0.14          -0.011              -1             1.8               0               0               0               0               0           -0.37            -0.3           -0.04               0               0        6e+03      0.092        1      0.6    +
    1            -1.3             1.1            0.11            -1.3             1.1               0               0               0               0               0            0.15          -0.018           -0.23               0               0      5.5e+03      0.018       10      1.1   ++
    2           -0.74             1.1             1.1            -2.2             0.1               0               0               0               0               0            0.56           -0.13           -0.84               0               0      5.5e+03      0.042       10     0.29    +
    3            -1.2             1.1             0.9            -1.6            0.28               0               0               0               0               0            0.24          -0.079           -0.78               0               0      5.5e+03     0.0026    1e+02        1   ++
    4            -1.2             1.1            0.88            -1.6            0.43               0               0               0               0               0            0.25          -0.081           -0.78               0               0      5.5e+03     0.0013    1e+03     0.95   ++
    5            -1.2             1.1            0.89            -1.5            0.42               0               0               0               0               0            0.24          -0.079           -0.77               0               0      5.5e+03    7.2e-06    1e+04        1   ++
    6            -1.2             1.1            0.89            -1.5            0.42               0               0               0               0               0            0.24          -0.079           -0.77               0               0      5.5e+03    1.1e-09    1e+04        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 14 unknown parameters [max: 50]
*** Estimate b07everything_000341
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time       lambda_tt          b_cost beta_TRAIN_COST beta_TRAIN_COST       b_headway beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0           -0.72            0.65           -0.95               2               0               0               0         -0.0015               0               0           -0.41            -0.3               0               0        6e+03        2.3        1     0.54    +
    1           -0.93             1.6           -0.38             1.9               0               0               0         -0.0073               0               0           -0.15           -0.73               0               0      5.5e+03       0.17        1     0.86    +
    2            -1.2             1.8           -0.83            0.89               0               0               0         -0.0047               0               0           -0.24           -0.85               0               0      5.3e+03      0.026       10        1   ++
    3           -0.51             1.8            -1.7           -0.05               0               0               0         -0.0059               0               0            0.19            -1.3               0               0      5.2e+03       0.12       10     0.49    +
    4           -0.51             1.9            -1.7            0.32               0               0               0          -0.006               0               0            0.18            -1.4               0               0      5.2e+03      0.015    1e+02        1   ++
    5            -0.6             1.9            -1.6            0.34               0               0               0          -0.006               0               0            0.12            -1.3               0               0      5.2e+03    0.00078    1e+03     0.98   ++
    6            -0.6             1.9            -1.6            0.34               0               0               0          -0.006               0               0            0.12            -1.3               0               0      5.2e+03    8.4e-07    1e+03        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 23 unknown parameters [max: 50]
*** Estimate b07everything_000342
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.8e+03        2.6       10        1   ++
    1      5.5e+03       0.94    1e+02        1   ++
    2      5.5e+03      0.089    1e+03      1.1   ++
    3      5.5e+03     0.0022    1e+04        1   ++
    4      5.5e+03    1.6e-06    1e+04        1   ++
Considering neighbor 2/20 for current solution
Considering neighbor 3/20 for current solution
Attempt 92/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000343
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_ma beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.67           -0.94               0               0               0              -1        -0.00071               0               0           -0.61           -0.21               0               0      5.6e+03        2.8       10      1.1   ++
    1           -0.53            -1.3               0               0               0            -2.1         -0.0048               0               0           -0.97            0.13               0               0      5.4e+03       0.29    1e+02      1.1   ++
    2           -0.49            -1.4               0               0               0            -2.2         -0.0058               0               0              -1            0.16               0               0      5.4e+03      0.012    1e+03        1   ++
    3           -0.49            -1.4               0               0               0            -2.2         -0.0059               0               0            -1.1            0.16               0               0      5.4e+03    1.9e-05    1e+04        1   ++
    4           -0.49            -1.4               0               0               0            -2.2         -0.0059               0               0            -1.1            0.16               0               0      5.4e+03    3.9e-11    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 93/100
Biogeme parameters read from biogeme.toml.
Model with 23 unknown parameters [max: 50]
*** Estimate b07everything_000344
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.7e+03          3       10        1   ++
    1      5.6e+03       0.41    1e+02      1.1   ++
    2      5.6e+03      0.034    1e+03      1.1   ++
    3      5.6e+03    0.00026    1e+04        1   ++
    4      5.6e+03    2.1e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 94/100
Biogeme parameters read from biogeme.toml.
Model with 13 unknown parameters [max: 50]
*** Estimate b07everything_000345
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time beta_TRAIN_TT_S beta_TRAIN_TT_S          b_cost       b_headway beta_SM_TT_SCAL beta_SM_TT_SCAL     asc_car_ref asc_car_diff_wi beta_CAR_TT_SCA beta_CAR_TT_SCA     Function    Relgrad   Radius      Rho
    0           -0.91            0.23               0               0               0           -0.42          -0.008               0               0              -1            -0.2               0               0      5.7e+03        2.5       10        1   ++
    1            -1.7             2.2               0               0               0           -0.97         -0.0062               0               0           -0.97          0.0081               0               0      5.3e+03        1.1    1e+02        1   ++
    2            -1.9             2.2               0               0               0              -1         -0.0064               0               0              -1          -0.049               0               0      5.3e+03      0.062    1e+03        1   ++
    3            -1.9             2.2               0               0               0              -1         -0.0064               0               0              -1          -0.042               0               0      5.3e+03    0.00046    1e+04        1   ++
    4            -1.9             2.2               0               0               0              -1         -0.0064               0               0              -1          -0.042               0               0      5.3e+03    2.6e-08    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 95/100
Biogeme parameters read from biogeme.toml.
Model with 18 unknown parameters [max: 50]
*** Estimate b07everything_000346
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0        6e+03      0.073       10      1.1   ++
    1      5.5e+03      0.031    1e+02     0.98   ++
    2      5.5e+03     0.0011    1e+03        1   ++
    3      5.5e+03    4.7e-06    1e+03        1   ++
Considering neighbor 0/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 10 unknown parameters [max: 50]
*** Estimate b07everything_000347
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_on asc_car_diff_se asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.96            0.24          -0.014            0.89           -0.79              -1           -0.16           -0.16          -0.054           -0.33      5.2e+03      0.043       10      1.1   ++
    1            -1.6            0.54            0.39             1.7              -1            -2.5           -0.36           0.088           -0.25            0.99        5e+03      0.015    1e+02      1.1   ++
    2            -1.8            0.69            0.56             1.8            -1.1            -2.8           -0.38             0.1           -0.31             1.2        5e+03     0.0012    1e+03        1   ++
    3            -1.8            0.71            0.58             1.8            -1.1            -2.8           -0.37             0.1           -0.31             1.2        5e+03    1.1e-05    1e+04        1   ++
    4            -1.8            0.71            0.58             1.8            -1.1            -2.8           -0.37             0.1           -0.31             1.2        5e+03    1.2e-09    1e+04        1   ++
Considering neighbor 1/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 12 unknown parameters [max: 50]
*** Estimate b07everything_000348
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost beta_TRAIN_COST beta_TRAIN_COST beta_SM_COST_SC beta_SM_COST_SC     asc_car_ref asc_car_diff_wi beta_CAR_CO_SCA beta_CAR_CO_SCA     Function    Relgrad   Radius      Rho
    0              -1           0.036              -1               0               0               0               0               0         -0.0065              -1               0               0      5.7e+03      0.038       10      1.1   ++
    1           -0.95             2.1            -2.6               0               0               0               0               0            0.14            -1.2               0               0      5.3e+03      0.027    1e+02     0.92   ++
    2           -0.91             1.8              -3               0               0               0               0               0            0.21            -1.4               0               0      5.2e+03    0.00075    1e+03        1   ++
    3           -0.91             1.8              -3               0               0               0               0               0            0.21            -1.4               0               0      5.2e+03    1.9e-06    1e+03        1   ++
Considering neighbor 2/20 for current solution
Biogeme parameters read from biogeme.toml.
Model with 6 unknown parameters [max: 50]
*** Estimate b07everything_000349
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1          -0.065              -1           -0.72          -0.048           -0.46      5.6e+03      0.042       10      1.1   ++
    1            -1.2             2.4            -2.5            -2.2           -0.17            0.87        5e+03      0.039    1e+02     0.94   ++
    2            -1.1             2.1              -3            -2.7           -0.16             1.1        5e+03     0.0016    1e+03        1   ++
    3            -1.1             2.1            -3.1            -2.8           -0.15             1.2        5e+03    1.6e-05    1e+04        1   ++
    4            -1.1             2.1            -3.1            -2.8           -0.15             1.2        5e+03    1.3e-09    1e+04        1   ++
Considering neighbor 3/20 for current solution
*** New pareto solution:
asc:GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:sqrt [9929.037358349122, np.float64(9969.957123756838), 6]
Attempt 96/100
Considering neighbor 0/20 for current solution
Attempt 97/100
Considering neighbor 0/20 for current solution
Attempt 98/100
Biogeme parameters read from biogeme.toml.
Model with 16 unknown parameters [max: 50]
*** Estimate b07everything_000350
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Analytical Hessian method: full
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.     Function    Relgrad   Radius      Rho
    0      5.9e+03       0.04       10        1   ++
    1      5.3e+03      0.056    1e+02        1   ++
    2      5.2e+03      0.004    1e+03        1   ++
    3      5.2e+03    0.00015    1e+04        1   ++
    4      5.2e+03    1.7e-07    1e+04        1   ++
Considering neighbor 0/20 for current solution
Considering neighbor 1/20 for current solution
Attempt 99/100
Considering neighbor 0/20 for current solution
Pareto file has been updated: b22_multiple_models.pareto
Before the algorithm: 1 models, with 1 Pareto.
After the algorithm: 166 models, with 7 Pareto.
VNS algorithm completed. Postprocessing of the Pareto optimal solutions
Pareto set initialized from file with 166 elements [7 Pareto] and 0 invalid elements.
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000000.iter
Cannot read file __b22_multiple_models_000000.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1          -0.065              -1           -0.72          -0.048           -0.46      5.6e+03      0.042       10      1.1   ++
    1            -1.2             2.4            -2.5            -2.2           -0.17            0.87        5e+03      0.039    1e+02     0.94   ++
    2            -1.1             2.1              -3            -2.7           -0.16             1.1        5e+03     0.0016    1e+03        1   ++
    3            -1.1             2.1            -3.1            -2.8           -0.15             1.2        5e+03    1.6e-05    1e+04        1   ++
    4            -1.1             2.1            -3.1            -2.8           -0.15             1.2        5e+03    1.3e-09    1e+04        1   ++
Optimization algorithm has converged.
Relative gradient: 1.3349627072966304e-09
Cause of termination: Relative gradient = 1.3e-09 <= 6.1e-06
Number of function evaluations: 16
Number of gradient evaluations: 11
Number of hessian evaluations: 5
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 5
Proportion of Hessian calculation: 5/5 = 100.0%
Optimization time: 0:00:02.813950
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000001.iter
Cannot read file __b22_multiple_models_000001.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.59           -0.77            0.75              -1           -0.86           -0.46           -0.18           -0.28      5.3e+03      0.047       10      1.1   ++
    1           -0.43           -0.97             1.8            -2.7            -2.5            -0.5            0.37            0.88      4.9e+03      0.019    1e+02      1.1   ++
    2           -0.37            -1.1               2            -3.1            -2.7           -0.54            0.44               1      4.8e+03     0.0014    1e+03        1   ++
    3           -0.36            -1.1               2            -3.1            -2.8           -0.54            0.44               1      4.8e+03    7.5e-06    1e+04        1   ++
    4           -0.36            -1.1               2            -3.1            -2.8           -0.54            0.44               1      4.8e+03    2.5e-10    1e+04        1   ++
Optimization algorithm has converged.
Relative gradient: 2.54955549725972e-10
Cause of termination: Relative gradient = 2.5e-10 <= 6.1e-06
Number of function evaluations: 16
Number of gradient evaluations: 11
Number of hessian evaluations: 5
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 5
Proportion of Hessian calculation: 5/5 = 100.0%
Optimization time: 0:00:02.052012
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000002.iter
Cannot read file __b22_multiple_models_000002.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042       10      1.1   ++
    1              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042        5 -1.4e+07    -
    2              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042      2.5 -2.1e+02    -
    3              -1            0.11              -1           -0.27             1.1          -0.047              -1      5.6e+03      0.042      1.2    -0.26    -
    4            -1.4             1.4            -1.9            -1.1            0.99           -0.38            -1.1      5.1e+03      0.023       12      1.1   ++
    5            -1.4             1.4            -1.9            -1.1            0.99           -0.38            -1.1      5.1e+03      0.023      1.2      -12    -
    6            -1.3             2.5              -3              -2           -0.19           -0.31            -1.4        5e+03      0.041      1.2      0.5    +
    7            -1.3             2.1            -2.9            -1.4           -0.13           -0.18            -1.8        5e+03     0.0015       12     0.98   ++
    8            -1.2             2.1              -3            -1.5             0.1           -0.19            -1.9      4.9e+03      0.002  1.2e+02     0.91   ++
    9            -1.2             2.1              -3            -1.5           0.091           -0.19            -1.9      4.9e+03    1.9e-05  1.2e+03        1   ++
   10            -1.2             2.1              -3            -1.5           0.091           -0.19            -1.9      4.9e+03    2.3e-08  1.2e+03        1   ++
Optimization algorithm has converged.
Relative gradient: 2.3131979598718067e-08
Cause of termination: Relative gradient = 2.3e-08 <= 6.1e-06
Number of function evaluations: 26
Number of gradient evaluations: 15
Number of hessian evaluations: 7
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 11
Proportion of Hessian calculation: 7/7 = 100.0%
Optimization time: 0:00:02.428096
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000003.iter
Cannot read file __b22_multiple_models_000003.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.79           -0.72              -1           -0.42      5.5e+03      0.037       10      1.1   ++
    1           -0.75            -1.2            -2.2            -0.1      5.3e+03     0.0076    1e+02      1.1   ++
    2           -0.73            -1.2            -2.3           -0.11      5.3e+03    0.00023    1e+03        1   ++
    3           -0.73            -1.2            -2.3           -0.11      5.3e+03    2.7e-07    1e+03        1   ++
Optimization algorithm has converged.
Relative gradient: 2.7148828825437547e-07
Cause of termination: Relative gradient = 2.7e-07 <= 6.1e-06
Number of function evaluations: 13
Number of gradient evaluations: 9
Number of hessian evaluations: 4
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 4
Proportion of Hessian calculation: 4/4 = 100.0%
Optimization time: 0:00:01.447006
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000004.iter
Cannot read file __b22_multiple_models_000004.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost       b_headway     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.47            -0.7            0.44              -1           -0.64               1          0.0022           -0.28          -0.067           -0.19      5.3e+03        2.7       10      1.1   ++
    1           -0.47            -0.7            0.44              -1           -0.64               1          0.0022           -0.28          -0.067           -0.19      5.3e+03        2.7      1.3     -3.2    -
    2           -0.12              -1             1.7            -1.4            -1.2            0.84         -0.0024           -0.33            0.21           -0.62      4.9e+03        0.7       13        1   ++
    3           -0.12              -1             1.7            -1.4            -1.2            0.84         -0.0024           -0.33            0.21           -0.62      4.9e+03        0.7     0.64    0.017    -
    4           0.019            -1.2             2.2            -1.6            -1.4             0.2         -0.0063           -0.52            0.31            -1.1      4.8e+03       0.15      6.4      1.1   ++
    5          -0.071            -1.2             2.2            -1.6            -1.5          -0.045         -0.0066           -0.68            0.45            -1.8      4.8e+03       0.03       64        1   ++
    6          -0.071            -1.2             2.2            -1.6            -1.5          -0.036         -0.0066           -0.69            0.46              -2      4.8e+03     0.0014  6.4e+02        1   ++
    7          -0.071            -1.2             2.2            -1.6            -1.5          -0.036         -0.0066           -0.69            0.46              -2      4.8e+03    4.6e-06  6.4e+02        1   ++
Optimization algorithm has converged.
Relative gradient: 4.589403692686364e-06
Cause of termination: Relative gradient = 4.6e-06 <= 6.1e-06
Number of function evaluations: 21
Number of gradient evaluations: 13
Number of hessian evaluations: 6
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 8
Proportion of Hessian calculation: 6/6 = 100.0%
Optimization time: 0:00:03.526943
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000005.iter
Cannot read file __b22_multiple_models_000005.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.       asc_train          b_time       lambda_tt          b_cost         asc_car     Function    Relgrad   Radius      Rho
    0           -0.68              -1             1.9           -0.64           -0.59        6e+03       0.11        1     0.59    +
    1           -0.92           -0.75             1.6            -1.6          -0.024      5.4e+03      0.031       10     0.96   ++
    2           -0.92           -0.75             1.6            -1.6          -0.024      5.4e+03      0.031      1.2     -1.7    -
    3           -0.68            -1.8            0.39            -2.3            0.03      5.3e+03       0.04      1.2     0.88    +
    4           -0.48            -1.7            0.47            -2.3           0.054      5.2e+03     0.0029       12     0.95   ++
    5            -0.5            -1.7            0.48            -2.4           0.057      5.2e+03    2.1e-05  1.2e+02        1   ++
    6            -0.5            -1.7            0.48            -2.4           0.057      5.2e+03    1.2e-09  1.2e+02        1   ++
Optimization algorithm has converged.
Relative gradient: 1.240651099909684e-09
Cause of termination: Relative gradient = 1.2e-09 <= 6.1e-06
Number of function evaluations: 20
Number of gradient evaluations: 13
Number of hessian evaluations: 6
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 7
Proportion of Hessian calculation: 6/6 = 100.0%
Optimization time: 0:00:02.718280
Calculate final gradient and BHHH
Calculate second derivatives
Biogeme parameters provided by the user.
*** Initial values of the parameters are obtained from the file __b22_multiple_models_000006.iter
Cannot read file __b22_multiple_models_000006.iter. Statement is ignored.
Starting values for the algorithm: {}
Analytical Hessian method: full
As the model is not too complex, we activate the calculation of second derivatives. To change this behavior, modify the algorithm to "simple_bounds" in the TOML file.
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: Newton with trust region for simple bounds
Iter.   asc_train_ref asc_train_diff_ asc_train_diff_          b_time          b_cost     lambda_cost     asc_car_ref asc_car_diff_ma asc_car_diff_wi     Function    Relgrad   Radius      Rho
    0           -0.43           -0.58            0.47              -1           -0.62               1           -0.37           -0.19           -0.21      5.3e+03       0.05       10        1   ++
    1           -0.43           -0.58            0.47              -1           -0.62               1           -0.37           -0.19           -0.21      5.3e+03       0.05      1.3       -3    -
    2           -0.32            -1.1             1.8            -1.6            -1.1            0.81           -0.38            0.15           -0.66      4.9e+03      0.019       13      1.1   ++
    3           -0.32            -1.1             1.8            -1.6            -1.1            0.81           -0.38            0.15           -0.66      4.9e+03      0.019     0.82     -2.4    -
    4           -0.26            -1.2             2.2            -1.7            -1.3          -0.011           -0.37            0.38            -1.1      4.8e+03     0.0087      8.2     0.99   ++
    5           -0.38            -1.2             2.1            -1.6            -1.5          -0.026           -0.55            0.46            -1.8      4.8e+03     0.0014       82      1.1   ++
    6           -0.38            -1.2             2.1            -1.6            -1.5           -0.04           -0.56            0.46              -2      4.8e+03    8.5e-05  8.2e+02        1   ++
    7           -0.38            -1.2             2.1            -1.6            -1.5           -0.04           -0.56            0.46              -2      4.8e+03    2.8e-07  8.2e+02        1   ++
Optimization algorithm has converged.
Relative gradient: 2.8139970693199444e-07
Cause of termination: Relative gradient = 2.8e-07 <= 6.1e-06
Number of function evaluations: 21
Number of gradient evaluations: 13
Number of hessian evaluations: 6
Algorithm: Newton with trust region for simple bound constraints
Number of iterations: 8
Proportion of Hessian calculation: 6/6 = 100.0%
Optimization time: 0:00:02.997571
Calculate final gradient and BHHH
Calculate second derivatives
Pareto: 7
Considered: 166
Removed: 12
summary, description = compile_estimation_results(
    non_dominated_models, use_short_names=True
)
print(summary)
                                    Model_000000  ...       Model_000006
Number of estimated parameters                 6  ...                  9
Sample size                                 6768  ...               6768
Final log likelihood                   -4958.519  ...          -4830.135
Akaike Information Criterion            9929.037  ...            9678.27
Bayesian Information Criterion          9969.957  ...           9739.649
asc_train_ref (t-test)              -1.14  (-14)  ...    -0.382  (-4.12)
asc_train_diff_with_ga (t-test)     2.07  (23.7)  ...       2.13  (23.5)
b_time (t-test)                     -3.11  (-17)  ...     -1.62  (-20.1)
b_cost (t-test)                   -2.79  (-17.2)  ...      -1.5  (-18.3)
asc_car_ref (t-test)             -0.153  (-2.64)  ...    -0.557  (-5.27)
asc_car_diff_with_ga (t-test)       1.15  (5.02)  ...     -2.02  (-9.06)
asc_train_diff_male (t-test)                      ...     -1.17  (-13.8)
asc_car_diff_male (t-test)                        ...      0.465  (4.28)
lambda_cost (t-test)                              ...  -0.0401  (-0.444)
asc_train (t-test)                                ...
asc_car (t-test)                                  ...
b_headway (t-test)                                ...
lambda_tt (t-test)                                ...

[18 rows x 7 columns]

Explanation of the short names of the model.

for k, v in description.items():
    if k != v:
        print(f'{k}: {v}')
Model_000000: asc:GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:sqrt
Model_000001: asc:MALE-GA;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:sqrt
Model_000002: asc:GA;train_cost_catalog:boxcox;train_headway_catalog:without_headway;train_tt_catalog:sqrt
Model_000003: asc:no_seg;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:linear
Model_000004: asc:MALE-GA;train_cost_catalog:boxcox;train_headway_catalog:with_headway;train_tt_catalog:log
Model_000005: asc:no_seg;train_cost_catalog:sqrt;train_headway_catalog:without_headway;train_tt_catalog:boxcox
Model_000006: asc:MALE-GA;train_cost_catalog:boxcox;train_headway_catalog:without_headway;train_tt_catalog:log

Total running time of the script: (7 minutes 37.310 seconds)

Gallery generated by Sphinx-Gallery