def checkDerivatives(theFunction, x, names=None, logg=False):
"""Verifies the analytical derivatives of a function by comparing them with finite
difference approximations.
:param theFunction: A function object that takes a vector as an argument, and
returns a tuple.
- The first element of the tuple is the value of the function :math:`f`,
- the second is the gradient of the function,
- the third is the hessian.
:type theFunction: function
:param x: arguments of the function
:type x: numpy.array
:param names: the names of the entries of x (for reporting).
:type names: list(string)
:param logg: if True, messages will be displayed.
:type logg: bool
:return: tuple f, g, h, gdiff, hdiff where
- f is the value of the function at x,
- g is the analytical gradient,
- h is the analytical hessian,
- gdiff is the difference between the analytical gradient and the finite
difference approximation
- hdiff is the difference between the analytical hessian and the finite
difference approximation
:rtype: float, numpy.array,numpy.array, numpy.array,numpy.array
"""
f, g, h = theFunction(x)
g_num = findiff_g(theFunction, x)
gdiff = g - g_num
if logg:
logger = msg.bioMessage()
if names is None:
names = [f'x[{i}]' for i in range(len(x))]
logger.detailed('x\t\tGradient\tFinDiff\t\tDifference')
for k, v in enumerate(gdiff):
logger.detailed(f'{names[k]:15}\t{g[k]:+E}\t{g_num[k]:+E}\t{v:+E}')
h_num = findiff_H(theFunction, x)
hdiff = h - h_num
if logg:
logger.detailed('Row\t\tCol\t\tHessian\tFinDiff\t\tDifference')
for row in range(len(hdiff)):
for col in range(len(hdiff)):
logger.detailed(
f'{names[row]:15}\t{names[col]:15}\t{h[row,col]:+E}\t'
f'{h_num[row,col]:+E}\t{hdiff[row,col]:+E}')
return f, g, h, gdiff, hdiff
B_COST * SM_COST_SCALED
V3 = ASC_CAR + \
B_TIME * models.boxcox(CAR_TT_SCALED, LAMBDA) + \
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}
# Definition of the model. This is the contribution of each
# observation to the log likelihood function.
logprob = models.loglogit(V, av, CHOICE)
# Define level of verbosity
logger = msg.bioMessage()
logger.setSilent()
#logger.setWarning()
#logger.setGeneral()
#logger.setDetailed()
# Create the Biogeme object
biogeme = bio.BIOGEME(database, logprob)
biogeme.modelName = '08boxcox'
# Estimate the parameters
results = biogeme.estimate()
pandasResults = results.getEstimatedParameters()
print(pandasResults)
def __init__(self, name, pandasDatabase):
"""Constructor
:param name: name of the database.
:type name: string
:param pandasDatabase: data stored in a pandas data frame.
:type pandasDatabase: pandas.DataFrame
"""
self.logger = msg.bioMessage()
start_time = datetime.now()
## Name of the database. Used mainly for the file name when dumping data.
self.name = name
## Pandas data frame containing the data.
self.data = pandasDatabase
self.fullData = pandasDatabase
## self.variables is initialized by _generateHeaders()
self.variables = None
self._generateHeaders()
## Number of observations removed by the function Database.remove
self.excludedData = 0
## Name of the column identifying the individuals in a panel
## data context. None if data is not panel.
self.panelColumn = None
## map identifying the range of observations for each
## individual in a panel data context. None if data is not
## panel.
self.individualMap = None
self.fullIndividualMap = None
## Initialize the dictionary containing random number
## generators with a series of native generators.
self._initNativeRandomNumberGenerators()
## Dictionary containing user defined random number
## generators. Defined by the function
## Database.setRandomNumberGenerators that checks that
## reserved keywords are not used. The element of the
## dictionary is a tuple with two elements: (0) the function
## generating the draws, and (1) a string describing the type of draws
self.userRandomNumberGenerators = dict()
## Number of draws generated by the function Database.generateDraws.
## Value 0 if this function is not called.
self.numberOfDraws = 0
## Types of draws for Monte Carlo integration
self.typesOfDraws = {}
self._auditDone = False
## Draws for Monte-Carlo integration
self.theDraws = None
## Availability expression to check
self._avail = None
## Choice expression to check
self._choice = None
## Expression to check
self._expression = None
listOfErrors, listOfWarnings = self._audit()
if listOfWarnings:
self.logger.warning('\n'.join(listOfWarnings))
if listOfErrors:
self.logger.warning('\n'.join(listOfErrors))
raise excep.biogemeError('\n'.join(listOfErrors))
def run_simulation(data_file_directory_for_simulation, data_file_name_for_simulation, output_directory_for_simulation,
betas, household_income_limit):
"""
:author: Antonin Danalet, based on the example '01logit_simul.py' by Michel Bierlaire, EPFL, on biogeme.epfl.ch
Simulation with a binary logit model. Two alternatives: work from home at least some times, or not."""
# Read the data
df_persons = pd.read_csv(data_file_directory_for_simulation / data_file_name_for_simulation, ';')
database = db.Database('persons', df_persons)
# The following statement allows you to use the names of the variable as Python variable.
globals().update(database.variables)
# Parameters to be estimated
alternative_specific_constant = Beta('alternative_specific_constant', 0, None, None, 0)
b_no_post_school_education = Beta('b_no_post_school_education', 0, None, None, 0)
b_secondary_education = Beta('b_secondary_education', 0, None, None, 0)
b_tertiary_education = Beta('b_tertiary_education', 0, None, None, 0)
b_university = Beta('b_university', 0, None, None, 1)
b_male = Beta('b_male', 0, None, None, 0)
b_public_transport_connection_quality_na_home = Beta('b_public_transport_connection_quality_na_home',
0, None, None, 0)
b_public_transport_connection_quality_a_work = Beta('b_public_transport_connection_quality_are_a_work',
0, None, None, 1)
b_rural_work = Beta('b_rural_work', 0, None, None, 0)
b_home_work_distance = Beta('b_home_work_distance', 0, None, None, 0)
b_business_sector_agriculture = Beta('b_business_sector_agriculture', 0, None, None, 0)
b_business_sector_production = Beta('b_business_sector_production', 0, None, None, 0)
b_business_sector_wholesale = Beta('b_business_sector_wholesale', 0, None, None, 1)
b_business_sector_retail = Beta('b_business_sector_retail', 0, None, None, 0)
b_business_sector_gastronomy = Beta('b_business_sector_gastronomy', 0, None, None, 0)
b_business_sector_finance = Beta('b_business_sector_finance', 0, None, None, 1)
b_business_sector_services_fc = Beta('b_business_sector_services_fc', 0, None, None, 0)
b_business_sector_other_services = Beta('b_business_sector_other_services', 0, None, None, 1)
b_business_sector_others = Beta('b_business_sector_others', 0, None, None, 1)
b_business_sector_non_movers = Beta('b_business_sector_non_movers', 0, None, None, 0)
b_executives = Beta('b_executives', 0, None, None, 0)
b_german = Beta('b_german', 0, None, None, 0)
b_hh_income_8000_or_less = Beta('b_hh_income_8000_or_less', 0, None, None, 0)
# Definition of new variables
no_post_school_educ = education == 1
secondary_education = education == 2
tertiary_education = education == 3
university = education == 4
male = (sex == 1)
public_transport_quality_NA_home = (public_transport_connection_quality_ARE_home == 5)
public_transport_quality_A_work = (public_transport_connection_quality_ARE_work == 1)
home_work_distance = (home_work_crow_fly_distance * (home_work_crow_fly_distance >= 0.0) / 100000.0)
business_sector_agriculture = type_1 == 1
business_sector_retail = type_1 == 4
business_sector_gastronomy = type_1 == 5
business_sector_finance = type_1 == 6
business_sector_production = type_1 == 2
business_sector_wholesale = type_1 == 3
business_sector_services_fC = type_1 == 7
business_sector_other_services = type_1 == 8
business_sector_others = type_1 == 9
business_sector_non_movers = type_1 == 10
german = language == 1
nationality_switzerland = nation == 0
nationality_germany_austria = nation == 1
nationality_italy_vatican = nation == 2
nationality_france_monaco_s_marino = nation == 3
nationality_northwestern_europe = nation == 4
nationality_eastern_europe = nation == 7
hh_income_8000_or_less = hh_income < household_income_limit
executives = (0 < position_in_bus) * (position_in_bus < 19)
rural_work = urban_rural_typology_work == 3
# Utility
utility_function_telecommuting = alternative_specific_constant + \
b_executives * executives + \
b_no_post_school_education * no_post_school_educ + \
b_secondary_education * secondary_education + \
b_tertiary_education * tertiary_education + \
b_university * university + \
b_male * male + \
b_public_transport_connection_quality_na_home * public_transport_quality_NA_home + \
b_public_transport_connection_quality_a_work * public_transport_quality_A_work + \
b_rural_work * rural_work + \
b_home_work_distance * home_work_distance + \
models.piecewiseFormula(age, [0, 20, 35, 75, 200]) + \
b_business_sector_agriculture * business_sector_agriculture + \
b_business_sector_retail * business_sector_retail + \
b_business_sector_gastronomy * business_sector_gastronomy + \
b_business_sector_finance * business_sector_finance + \
b_business_sector_production * business_sector_production + \
b_business_sector_wholesale * business_sector_wholesale + \
b_business_sector_services_fc * business_sector_services_fC + \
b_business_sector_other_services * business_sector_other_services + \
b_business_sector_others * business_sector_others + \
b_business_sector_non_movers * business_sector_non_movers + \
b_german * german + \
b_nationality_ch_germany_france_italy_nw_e * nationality_switzerland + \
b_nationality_ch_germany_france_italy_nw_e * nationality_germany_austria + \
b_nationality_ch_germany_france_italy_nw_e * nationality_italy_vatican + \
b_nationality_ch_germany_france_italy_nw_e * nationality_france_monaco_s_marino + \
b_nationality_ch_germany_france_italy_nw_e * nationality_northwestern_europe + \
b_nationality_ch_germany_france_italy_nw_e * nationality_eastern_europe + \
models.piecewiseFormula(work_percentage, [0, 90, 101]) + \
b_hh_income_8000_or_less * hh_income_8000_or_less
utility_function_no_telecommuting = 0
# Associate utility functions with the numbering of alternatives
utility_functions_with_numbering_of_alternatives = {1: utility_function_telecommuting, # Yes or sometimes
3: utility_function_no_telecommuting} # No
availability_conditions = {1: 1, # Always available
3: 1} # Always available
# The choice model is a logit, with availability conditions
prob_telecommuting = models.logit(utility_functions_with_numbering_of_alternatives, availability_conditions, 1)
prob_no_telecommuting = models.logit(utility_functions_with_numbering_of_alternatives, availability_conditions, 3)
simulate = {'Prob. telecommuting': prob_telecommuting,
'Prob. no telecommuting': prob_no_telecommuting}
# Create the Biogeme object
biogeme = bio.BIOGEME(database, simulate)
biogeme.modelName = 'logit_telecommuting_simul'
# Define level of verbosity
logger = msg.bioMessage()
# logger.setSilent()
logger.setWarning()
# logger.setGeneral()
# logger.setDetailed()
# Get the betas from the estimation (without corrections)
# path_to_estimation_folder = Path('../data/output/models/estimation/')
# if os.path.isfile(path_to_estimation_folder / 'logit_telecommuting~00.pickle'):
# raise Exception('There are several model outputs! Careful.')
# results = res.bioResults(pickleFile=path_to_estimation_folder / 'logit_telecommuting.pickle')
# betas_without_correction = results.getBetaValues()
# Change the working directory, so that biogeme writes in the correct folder, i.e., where this file is
standard_directory = os.getcwd()
os.chdir(output_directory_for_simulation)
results = biogeme.simulate(theBetaValues=betas)
# print(results.describe())
df_persons = pd.concat([df_persons, results], axis=1)
# Go back to the normal working directory
os.chdir(standard_directory)
# For unemployed people, fix probability of doing some home office to 0 (and probability of not doing to 1).
df_persons.loc[df_persons.employed == 0, 'Prob. telecommuting'] = 0.0 # Unemployed people
df_persons.loc[df_persons.employed == 0, 'Prob. no telecommuting'] = 1.0 # Unemployed people
df_persons.loc[df_persons.employed == -99, 'Prob. telecommuting'] = 0.0 # Other people
df_persons.loc[df_persons.employed == -99, 'Prob. no telecommuting'] = 1.0 # Other people
# By definition, apprentices don't work from home (because they were not asked in the MTMC)
df_persons.loc[df_persons.position_in_bus == 3, 'Prob. telecommuting'] = 0.0
df_persons.loc[df_persons.position_in_bus == 3, 'Prob. no telecommuting'] = 1.0
# Add a realisation of the probability
df_persons['random 0/1'] = np.random.rand(len(df_persons))
df_persons['telecommuting_model'] = np.where(df_persons['random 0/1'] < df_persons['Prob. telecommuting'], 1, 0)
del df_persons['random 0/1']
''' Save the file '''
data_file_name = 'persons_from_SynPop_with_probability_telecommuting.csv'
df_persons.to_csv(output_directory_for_simulation / data_file_name, sep=',', index=False)
