Note
Go to the end to download the full example code.
Mixture of logit with Halton draws
Example of a mixture of logit models, using quasi Monte-Carlo integration with Halton draws (base 5). The mixing distribution is normal.
- author:
Michel Bierlaire, EPFL
- date:
Wed Apr 12 18:21:13 2023
import biogeme.biogeme_logging as blog
import biogeme.biogeme as bio
from biogeme import models
from biogeme.expressions import Beta, bioDraws, MonteCarlo, log
from biogeme.parameters import Parameters
See the data processing script: Data preparation for Swissmetro.
from swissmetro_data import (
database,
CHOICE,
CAR_AV_SP,
TRAIN_AV_SP,
TRAIN_TT_SCALED,
TRAIN_COST_SCALED,
SM_TT_SCALED,
SM_COST_SCALED,
CAR_TT_SCALED,
CAR_CO_SCALED,
SM_AV,
)
logger = blog.get_screen_logger(level=blog.INFO)
logger.info('Example b24halton_mixture.py')
Example b24halton_mixture.py
Parameters to be estimated.
ASC_CAR = Beta('ASC_CAR', 0, None, None, 0)
ASC_TRAIN = Beta('ASC_TRAIN', 0, None, None, 0)
ASC_SM = Beta('ASC_SM', 0, None, None, 1)
B_COST = Beta('B_COST', 0, None, None, 0)
Define a random parameter, normally distributed, designed to be used for Monte-Carlo simulation.
B_TIME = Beta('B_TIME', 0, None, None, 0)
It is advised not to use 0 as starting value for the following parameter.
B_TIME_S = Beta('B_TIME_S', 1, None, None, 0)
Define a random parameter with a normal distribution, designed to be used for quasi Monte-Carlo simulation with Halton draws (base 5).
B_TIME_RND = B_TIME + B_TIME_S * bioDraws('b_time_rnd', 'NORMAL_HALTON5')
Definition of the utility functions.
V1 = ASC_TRAIN + B_TIME_RND * TRAIN_TT_SCALED + B_COST * TRAIN_COST_SCALED
V2 = ASC_SM + B_TIME_RND * SM_TT_SCALED + B_COST * SM_COST_SCALED
V3 = ASC_CAR + B_TIME_RND * CAR_TT_SCALED + B_COST * CAR_CO_SCALED
Associate utility functions with the numbering of alternatives.
V = {1: V1, 2: V2, 3: V3}
Associate the availability conditions with the alternatives.
av = {1: TRAIN_AV_SP, 2: SM_AV, 3: CAR_AV_SP}
Conditional on b_time_rnd, we have a logit model (called the kernel)
prob = models.logit(V, av, CHOICE)
We integrate over b_time_rnd using Monte-Carlo.
logprob = log(MonteCarlo(prob))
These notes will be included as such in the report file.
USER_NOTES = (
'Example of a mixture of logit models with three alternatives, '
'approximated using Monte-Carlo integration with Halton draws.'
)
As the objective is to illustrate the syntax, we calculate the Monte-Carlo approximation with a small number of draws.
the_biogeme = bio.BIOGEME(
database, logprob, user_notes=USER_NOTES, number_of_draws=100, seed=1223
)
the_biogeme.modelName = 'b24halton_mixture'
Biogeme parameters read from biogeme.toml.
Estimate the parameters
results = the_biogeme.estimate()
As the model is rather complex, we cancel the calculation of second derivatives. If you want to control the parameters, change the name of the algorithm in the TOML file from "automatic" to "simple_bounds"
*** Initial values of the parameters are obtained from the file __b24halton_mixture.iter
Cannot read file __b24halton_mixture.iter. Statement is ignored.
The number of draws (100) is low. The results may not be meaningful.
As the model is rather complex, we cancel the calculation of second derivatives. If you want to control the parameters, change the name of the algorithm in the TOML file from "automatic" to "simple_bounds"
Optimization algorithm: hybrid Newton/BFGS with simple bounds [simple_bounds]
** Optimization: BFGS with trust region for simple bounds
Iter. ASC_CAR ASC_TRAIN B_COST B_TIME B_TIME_S Function Relgrad Radius Rho
0 1 -1 -1 -1 2 6.1e+03 0.16 1 0.25 +
1 0 -0.73 -0.4 -2 3 5.5e+03 0.049 1 0.36 +
2 0.51 -0.94 -1.4 -2.3 2.6 5.4e+03 0.054 1 0.39 +
3 0.51 -0.94 -1.4 -2.3 2.6 5.4e+03 0.054 0.5 -0.15 -
4 0.0062 -0.44 -1.1 -2.8 2.6 5.3e+03 0.03 0.5 0.5 +
5 0.33 -0.089 -1.6 -2.6 2.5 5.3e+03 0.046 0.5 0.13 +
6 0.33 -0.089 -1.6 -2.6 2.5 5.3e+03 0.046 0.25 -0.18 -
7 0.26 -0.34 -1.4 -2.9 2.3 5.2e+03 0.022 0.25 0.65 +
8 0.23 -0.3 -1.2 -2.6 2.2 5.2e+03 0.0087 0.25 0.52 +
9 0.23 -0.3 -1.2 -2.6 2.2 5.2e+03 0.0087 0.12 -2.2 -
10 0.23 -0.3 -1.2 -2.6 2.2 5.2e+03 0.0087 0.062 -0.13 -
11 0.29 -0.36 -1.3 -2.6 2.1 5.2e+03 0.0072 0.062 0.31 +
12 0.23 -0.3 -1.4 -2.6 2.1 5.2e+03 0.0065 0.062 0.58 +
13 0.21 -0.36 -1.3 -2.6 2 5.2e+03 0.011 0.062 0.32 +
14 0.22 -0.33 -1.3 -2.5 2 5.2e+03 0.0034 0.62 0.94 ++
15 0.22 -0.33 -1.3 -2.5 2 5.2e+03 0.0034 0.31 -0.026 -
16 0.13 -0.39 -1.3 -2.3 1.7 5.2e+03 0.0039 0.31 0.58 +
17 0.13 -0.39 -1.3 -2.3 1.7 5.2e+03 0.0039 0.16 -6.3 -
18 0.13 -0.39 -1.3 -2.3 1.7 5.2e+03 0.0039 0.078 -4.7 -
19 0.13 -0.39 -1.3 -2.3 1.7 5.2e+03 0.0039 0.039 -2.2 -
20 0.13 -0.39 -1.3 -2.3 1.7 5.2e+03 0.0039 0.02 -0.15 -
21 0.15 -0.41 -1.3 -2.3 1.7 5.2e+03 0.00097 0.02 0.41 +
22 0.14 -0.39 -1.3 -2.3 1.7 5.2e+03 0.00075 0.02 0.29 +
23 0.15 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00083 0.02 0.41 +
24 0.15 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00083 0.0098 -0.089 -
25 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00036 0.0098 0.46 +
26 0.14 -0.41 -1.3 -2.3 1.7 5.2e+03 0.00064 0.0098 0.28 +
27 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00031 0.0098 0.28 +
28 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00031 0.0049 -0.24 -
29 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00031 0.0024 -0.83 -
30 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00047 0.0024 0.1 +
31 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00023 0.024 0.94 ++
32 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00023 0.012 -8.4 -
33 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00023 0.0061 -2.8 -
34 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 0.00023 0.0031 -0.17 -
35 0.14 -0.4 -1.3 -2.3 1.7 5.2e+03 2.8e-05 0.0031 0.79 -
Results saved in file b24halton_mixture.html
Results saved in file b24halton_mixture.pickle
print(results.short_summary())
Results for model b24halton_mixture
Nbr of parameters: 5
Sample size: 6768
Excluded data: 3960
Final log likelihood: -5215.687
Akaike Information Criterion: 10441.37
Bayesian Information Criterion: 10475.47
pandas_results = results.get_estimated_parameters()
pandas_results
Total running time of the script: (0 minutes 12.833 seconds)