201: Exampville Mode Choice
Contents
201: Exampville Mode Choice¶
Welcome to Exampville, the best simulated town in this here part of the internet!
Exampville is a demonstration provided with Larch that walks through some of the data and tools that a transportation planner might use when building a travel model.
import os
import numpy as np
import pandas as pd
import larch.numba as lx
from larch import P, X
In this example notebook, we will walk through the creation of a tour mode choice model.
To begin, we’ll re-load the tours and skims data from the data setup example.
hh, pp, tour, skims = lx.example(200, ['hh', 'pp', 'tour', 'skims'])
Preprocessing¶
The Exampville data output contains a set of files similar to what we might find for a real travel survey: network skims, and tables of households, persons, and tours. We’ll need to connect these tables together to create a composite dataset for mode choice model estimation, using the DataTree structure.
from addicty import Dict
Mode = Dict(
DA = 1,
SR = 2,
Walk = 3,
Bike = 4,
Transit = 5,
).freeze()
tour_dataset = lx.Dataset.from_idco(tour.set_index('TOURID'), alts=Mode)
od_skims = lx.Dataset.from_omx(skims)
dt = lx.DataTree(
tour=tour_dataset,
hh=hh.set_index('HHID'),
person=pp.set_index('PERSONID'),
od=od_skims,
do=od_skims,
relationships=(
"tours.HHID @ hh.HHID",
"tours.PERSONID @ person.PERSONID",
"hh.HOMETAZ @ od.otaz",
"tours.DTAZ @ od.dtaz",
"hh.HOMETAZ @ do.dtaz",
"tours.DTAZ @ do.otaz",
),
)
In Exampville, there are only two kinds of trips:
work (purpose=1) and
non-work (purpose=2).
We want to estimate a mode choice model for work trips, so we’ll begin by excluding all the other trips:
dt_work = dt.query_cases("TOURPURP == 1")
Model Definition¶
And then we are ready to create our model.
m = lx.Model(datatree = dt_work)
m.title = "Exampville Work Tour Mode Choice v1"
We will explicitly define the set of utility functions
we want to use. Because the DataFrames we are using to
serve data to this model contains exclusively idco format
data, we’ll use only the utility_co mapping to define
a unique utility function for each alternative.
m.utility_co[Mode.DA] = (
+ P.InVehTime * X.AUTO_TIME
+ P.Cost * X.AUTO_COST # dollars per mile
)
m.utility_co[Mode.SR] = (
+ P.ASC_SR
+ P.InVehTime * X.AUTO_TIME
+ P.Cost * (X.AUTO_COST * 0.5) # dollars per mile, half share
+ P("LogIncome:SR") * X("log(INCOME)")
)
m.utility_co[Mode.Walk] = (
+ P.ASC_Walk
+ P.NonMotorTime * X.WALK_TIME
+ P("LogIncome:Walk") * X("log(INCOME)")
)
m.utility_co[Mode.Bike] = (
+ P.ASC_Bike
+ P.NonMotorTime * X.BIKE_TIME
+ P("LogIncome:Bike") * X("log(INCOME)")
)
m.utility_co[Mode.Transit] = (
+ P.ASC_Transit
+ P.InVehTime * X.TRANSIT_IVTT
+ P.OutVehTime * X.TRANSIT_OVTT
+ P.Cost * X.TRANSIT_FARE
+ P("LogIncome:Transit") * X('log(INCOME)')
)
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Input In [8], in <module>
----> 1 m.utility_co[Mode.DA] = (
2 + P.InVehTime * X.AUTO_TIME
3 + P.Cost * X.AUTO_COST # dollars per mile
4 )
6 m.utility_co[Mode.SR] = (
7 + P.ASC_SR
8 + P.InVehTime * X.AUTO_TIME
9 + P.Cost * (X.AUTO_COST * 0.5) # dollars per mile, half share
10 + P("LogIncome:SR") * X("log(INCOME)")
11 )
13 m.utility_co[Mode.Walk] = (
14 + P.ASC_Walk
15 + P.NonMotorTime * X.WALK_TIME
16 + P("LogIncome:Walk") * X("log(INCOME)")
17 )
AttributeError: 'NoneType' object has no attribute 'DA'
To write a nested logit mode, we’ll attach some nesting nodes to the
model’s graph. Each new_node allows us to define the set of
codes for the child nodes (elemental alternatives, or lower level nests)
as well as giving the new nest a name and assigning a logsum parameter.
The return value of this method is the node code for the newly created
nest, which then can potenially be used as a child code when creating
a higher level nest. We do this below, adding the ‘Car’ nest into the
‘Motor’ nest.
Car = m.graph.new_node(parameter='Mu:Car', children=[Mode.DA, Mode.SR], name='Car')
NonMotor = m.graph.new_node(parameter='Mu:NonMotor', children=[Mode.Walk, Mode.Bike], name='NonMotor')
Motor = m.graph.new_node(parameter='Mu:Motor', children=[Car, Mode.Transit], name='Motor')
/home/runner/work/larch/larch/larch/larch/dataset.py:1148: UserWarning: no defined ALTID
warnings.warn("no defined ALTID")
/home/runner/work/larch/larch/larch/larch/numba/model.py:899: UserWarning: cannot initialize graph, must define alternatives somehow
warnings.warn('cannot initialize graph, must define alternatives somehow')
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [9], in <module>
----> 1 Car = m.graph.new_node(parameter='Mu:Car', children=[Mode.DA, Mode.SR], name='Car')
2 NonMotor = m.graph.new_node(parameter='Mu:NonMotor', children=[Mode.Walk, Mode.Bike], name='NonMotor')
3 Motor = m.graph.new_node(parameter='Mu:Motor', children=[Car, Mode.Transit], name='Motor')
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
Let’s visually check on the nesting structure.
m.graph
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [10], in <module>
----> 1 m.graph
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
The tour mode choice model’s choice variable is indicated by
the code value in ‘TOURMODE’, and this can be
defined for the model using choice_co_code.
m.choice_co_code = 'TOURMODE'
We can also give a dictionary of availability conditions based
on values in the idco data, using the availability_co_vars
attribute. Alternatives that are always available can be indicated
by setting the criterion to 1.
m.availability_co_vars = {
Mode.DA: 'AGE >= 16',
Mode.SR: 1,
Mode.Walk: 'WALK_TIME < 60',
Mode.Bike: 'BIKE_TIME < 60',
Mode.Transit: 'TRANSIT_FARE>0',
}
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Input In [12], in <module>
1 m.availability_co_vars = {
----> 2 Mode.DA: 'AGE >= 16',
3 Mode.SR: 1,
4 Mode.Walk: 'WALK_TIME < 60',
5 Mode.Bike: 'BIKE_TIME < 60',
6 Mode.Transit: 'TRANSIT_FARE>0',
7 }
AttributeError: 'NoneType' object has no attribute 'DA'
Then let’s prepare this data for estimation. Even though the
data is already in memory, the load_data method is used to
pre-process the data, extracting the required values, pre-computing
the values of fixed expressions, and assembling the results into
contiguous arrays suitable for computing the log likelihood values
efficiently.
Model Estimation¶
We can check on some important statistics of this loaded data even before we estimate the model.
m.choice_avail_summary()
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [13], in <module>
----> 1 m.choice_avail_summary()
File ~/work/larch/larch/larch/larch/numba/model.py:2059, in NumbaModel.choice_avail_summary(self)
2051 """
2052 Generate a summary of choice and availability statistics.
2053
(...)
2056 pandas.DataFrame
2057 """
2058 from ..dataset import choice_avail_summary
-> 2059 self.unmangle()
2060 graph = None if self.is_mnl() else self.graph
2061 return choice_avail_summary(
2062 self.dataset,
2063 graph,
2064 self.availability_co_vars,
2065 )
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
If we are satisfied with the statistics we see above, we can go ahead and estimate the model.
m.set_cap(20) # improves optimization stability
result = m.maximize_loglike()
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [15], in <module>
----> 1 m.set_cap(20) # improves optimization stability
2 result = m.maximize_loglike()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:622, in larch.model.parameter_frame.ParameterFrame.set_cap()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:195, in larch.model.parameter_frame.ParameterFrame.pf.__get__()
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
After we find the best fitting parameters, we can compute
some variance-covariance statistics, incuding standard errors of
the estimates and t statistics, using calculate_parameter_covariance.
m.calculate_parameter_covariance()
error in calculate_parameter_covariance
Traceback (most recent call last):
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 896, in graph
self.initialize_graph()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 966, in initialize_graph
super().initialize_graph(
File "larch/model/controller.pyx", line 1308, in larch.model.controller.Model5c.initialize_graph
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "larch/model/abstract_model.pyx", line 1065, in larch.model.abstract_model.AbstractChoiceModel.calculate_parameter_covariance
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1773, in d2_loglike
return super().d2_loglike(
File "larch/model/abstract_model.pyx", line 566, in larch.model.abstract_model.AbstractChoiceModel.d2_loglike
File "larch/model/abstract_model.pyx", line 417, in larch.model.abstract_model.AbstractChoiceModel.loglike3
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1706, in loglike2
result_arrays, penalty = self._loglike_runner(
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1339, in _loglike_runner
args = self.__prepare_for_compute(
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1172, in __prepare_for_compute
self.unmangle()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1118, in unmangle
self.reflow_data_arrays()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 977, in reflow_data_arrays
if self.graph is None:
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 900, in graph
raise RuntimeError('cannot initialize graph, must define alternatives somehow')
RuntimeError: cannot initialize graph, must define alternatives somehow
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [17], in <module>
----> 1 m.calculate_parameter_covariance()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:1230, in larch.model.abstract_model.AbstractChoiceModel.calculate_parameter_covariance()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:1065, in larch.model.abstract_model.AbstractChoiceModel.calculate_parameter_covariance()
File ~/work/larch/larch/larch/larch/numba/model.py:1773, in NumbaModel.d2_loglike(self, x, start_case, stop_case, step_case, leave_out, keep_only, subsample)
1762 def d2_loglike(
1763 self,
1764 x=None,
(...)
1771 subsample=-1,
1772 ):
-> 1773 return super().d2_loglike(
1774 x=x,
1775 start_case=start_case,
1776 stop_case=stop_case,
1777 step_case=step_case,
1778 leave_out=leave_out,
1779 keep_only=keep_only,
1780 subsample=subsample,
1781 )
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:566, in larch.model.abstract_model.AbstractChoiceModel.d2_loglike()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:417, in larch.model.abstract_model.AbstractChoiceModel.loglike3()
File ~/work/larch/larch/larch/larch/numba/model.py:1706, in NumbaModel.loglike2(self, x, start_case, stop_case, step_case, persist, leave_out, keep_only, subsample, return_series, probability_only)
1692 def loglike2(
1693 self,
1694 x=None,
(...)
1704 probability_only=False,
1705 ):
-> 1706 result_arrays, penalty = self._loglike_runner(
1707 x,
1708 start_case=start_case,
1709 stop_case=stop_case,
1710 step_case=step_case,
1711 return_gradient=True,
1712 )
1713 result = dictx(
1714 ll=result_arrays.loglike.sum() * self.weight_normalization,
1715 dll=result_arrays.d_loglike.sum(0) * self.weight_normalization,
1716 )
1717 if start_case is None and stop_case is None and step_case is None:
File ~/work/larch/larch/larch/larch/numba/model.py:1339, in NumbaModel._loglike_runner(self, x, only_utility, return_gradient, return_probability, return_bhhh, start_case, stop_case, step_case)
1327 def _loglike_runner(
1328 self,
1329 x=None,
(...)
1336 step_case=None,
1337 ):
1338 caseslice = slice(start_case, stop_case, step_case)
-> 1339 args = self.__prepare_for_compute(
1340 x,
1341 allow_missing_ch=return_probability or (only_utility>0),
1342 caseslice=caseslice,
1343 )
1344 args_flags = args + (np.asarray([
1345 only_utility,
1346 return_probability,
1347 return_gradient,
1348 return_bhhh,
1349 ], dtype=np.int8),)
1350 try:
File ~/work/larch/larch/larch/larch/numba/model.py:1172, in NumbaModel.__prepare_for_compute(self, x, allow_missing_ch, allow_missing_av, caseslice)
1170 if self._dataframes is not None and not self._dataframes.is_computational_ready(activate=True):
1171 raise ValueError('DataFrames is not computational-ready')
-> 1172 self.unmangle()
1173 if x is not None:
1174 self.set_values(x)
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
Then we can review the results in a variety of report tables.
m.parameter_summary()
error in parameter_summary
Traceback (most recent call last):
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 896, in graph
self.initialize_graph()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 966, in initialize_graph
super().initialize_graph(
File "larch/model/controller.pyx", line 1308, in larch.model.controller.Model5c.initialize_graph
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "larch/model/parameter_frame.pyx", line 848, in larch.model.parameter_frame.ParameterFrame.parameter_summary
File "larch/model/parameter_frame.pyx", line 661, in larch.model.parameter_frame.ParameterFrame.pfo
File "larch/model/parameter_frame.pyx", line 195, in larch.model.parameter_frame.ParameterFrame.pf.__get__
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1118, in unmangle
self.reflow_data_arrays()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 977, in reflow_data_arrays
if self.graph is None:
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 900, in graph
raise RuntimeError('cannot initialize graph, must define alternatives somehow')
RuntimeError: cannot initialize graph, must define alternatives somehow
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [18], in <module>
----> 1 m.parameter_summary()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:1008, in larch.model.parameter_frame.ParameterFrame.parameter_summary()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:848, in larch.model.parameter_frame.ParameterFrame.parameter_summary()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:661, in larch.model.parameter_frame.ParameterFrame.pfo()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:195, in larch.model.parameter_frame.ParameterFrame.pf.__get__()
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
m.estimation_statistics()
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [20], in <module>
----> 1 m.estimation_statistics()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:922, in larch.model.abstract_model.AbstractChoiceModel.estimation_statistics()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:736, in larch.model.abstract_model.AbstractChoiceModel.loglike_null()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:631, in larch.model.parameter_frame.ParameterFrame.get_values()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:195, in larch.model.parameter_frame.ParameterFrame.pf.__get__()
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
Save and Report Model¶
report = lx.Reporter(title=m.title)
report.append('# Parameter Summary')
report.append(m.parameter_summary())
report
error in parameter_summary
Traceback (most recent call last):
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 896, in graph
self.initialize_graph()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 966, in initialize_graph
super().initialize_graph(
File "larch/model/controller.pyx", line 1308, in larch.model.controller.Model5c.initialize_graph
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "larch/model/parameter_frame.pyx", line 848, in larch.model.parameter_frame.ParameterFrame.parameter_summary
File "larch/model/parameter_frame.pyx", line 661, in larch.model.parameter_frame.ParameterFrame.pfo
File "larch/model/parameter_frame.pyx", line 195, in larch.model.parameter_frame.ParameterFrame.pf.__get__
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 1118, in unmangle
self.reflow_data_arrays()
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 977, in reflow_data_arrays
if self.graph is None:
File "/home/runner/work/larch/larch/larch/larch/numba/model.py", line 900, in graph
raise RuntimeError('cannot initialize graph, must define alternatives somehow')
RuntimeError: cannot initialize graph, must define alternatives somehow
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [22], in <module>
1 report.append('# Parameter Summary')
----> 2 report.append(m.parameter_summary())
3 report
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:1008, in larch.model.parameter_frame.ParameterFrame.parameter_summary()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:848, in larch.model.parameter_frame.ParameterFrame.parameter_summary()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:661, in larch.model.parameter_frame.ParameterFrame.pfo()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:195, in larch.model.parameter_frame.ParameterFrame.pf.__get__()
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
report << "# Estimation Statistics" << m.estimation_statistics()
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [23], in <module>
----> 1 report << "# Estimation Statistics" << m.estimation_statistics()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:922, in larch.model.abstract_model.AbstractChoiceModel.estimation_statistics()
File ~/work/larch/larch/larch/larch/model/abstract_model.pyx:736, in larch.model.abstract_model.AbstractChoiceModel.loglike_null()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:631, in larch.model.parameter_frame.ParameterFrame.get_values()
File ~/work/larch/larch/larch/larch/model/parameter_frame.pyx:195, in larch.model.parameter_frame.ParameterFrame.pf.__get__()
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
report << "# Utility Functions" << m.utility_functions()
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
File ~/work/larch/larch/larch/larch/numba/model.py:896, in NumbaModel.graph(self)
895 try:
--> 896 self.initialize_graph()
897 except ValueError:
File ~/work/larch/larch/larch/larch/numba/model.py:966, in NumbaModel.initialize_graph(self, dataframes, alternative_codes, alternative_names, root_id)
961 alternative_names = get_coords_array(
962 'altname', 'altnames', 'alt_name', 'alt_names',
963 'alternative_name', 'alternative_names',
964 )
--> 966 super().initialize_graph(
967 dataframes=dataframes,
968 alternative_codes=alternative_codes,
969 alternative_names=alternative_names,
970 root_id=root_id,
971 )
File ~/work/larch/larch/larch/larch/model/controller.pyx:1308, in larch.model.controller.Model5c.initialize_graph()
ValueError: must define dataframes, dataservice, or give alternative_codes explicitly
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
Input In [24], in <module>
----> 1 report << "# Utility Functions" << m.utility_functions()
File ~/work/larch/larch/larch/larch/model/model.py:470, in Model.utility_functions(self, subset, resolve_parameters)
450 def utility_functions(self, subset=None, resolve_parameters=False):
451 """
452 Generate an XHTML output of the utility function(s).
453
(...)
468 xmle.Elem
469 """
--> 470 self.unmangle()
471 from xmle import Elem
472 x = Elem('div')
File ~/work/larch/larch/larch/larch/numba/model.py:1118, in NumbaModel.unmangle(self, force)
1116 super().unmangle(force=force)
1117 if self._dataset is None or force:
-> 1118 self.reflow_data_arrays()
1119 if self._fixed_arrays is None or force:
1120 self._rebuild_fixed_arrays()
File ~/work/larch/larch/larch/larch/numba/model.py:977, in NumbaModel.reflow_data_arrays(self)
973 def reflow_data_arrays(self):
974 """
975 Reload the internal data_arrays so they are consistent with the datatree.
976 """
--> 977 if self.graph is None:
978 self._data_arrays = None
979 return
File ~/work/larch/larch/larch/larch/numba/model.py:900, in NumbaModel.graph(self)
898 import warnings
899 warnings.warn('cannot initialize graph, must define alternatives somehow')
--> 900 raise RuntimeError('cannot initialize graph, must define alternatives somehow')
901 return self._graph
RuntimeError: cannot initialize graph, must define alternatives somehow
report.save(
'/tmp/exampville_mode_choice.html',
overwrite=True,
metadata=m,
)
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Input In [25], in <module>
----> 1 report.save(
2 '/tmp/exampville_mode_choice.html',
3 overwrite=True,
4 metadata=m,
5 )
File /usr/share/miniconda3/envs/development/lib/python3.9/site-packages/xmle/reporter.py:97, in Reporter.save(self, filename, overwrite, archive_dir, metadata)
94 self.renumber_numbered_items()
96 from .xhtml import XHTML
---> 97 with XHTML(
98 filename,
99 overwrite=overwrite,
100 metadata=metadata,
101 archive_dir=archive_dir,
102 title=self._short_title,
103 ) as f:
104 f << self
105 return f._filename
File /usr/share/miniconda3/envs/development/lib/python3.9/site-packages/xmle/xhtml.py:335, in XHTML.__init__(self, filename, overwrite, archive_dir, title, css, extra_css, jquery, jqueryui, floating_tablehead, metadata, toc, favicon)
333 except ImportError:
334 import pickle as cloudpickle
--> 335 _meta = (base64.standard_b64encode(zlib.compress(cloudpickle.dumps(metadata)))).decode()
336 self.head << Elem(tag="meta", name='pythonmetadata', content=_meta)
File /usr/share/miniconda3/envs/development/lib/python3.9/site-packages/cloudpickle/cloudpickle_fast.py:73, in dumps(obj, protocol, buffer_callback)
69 with io.BytesIO() as file:
70 cp = CloudPickler(
71 file, protocol=protocol, buffer_callback=buffer_callback
72 )
---> 73 cp.dump(obj)
74 return file.getvalue()
File /usr/share/miniconda3/envs/development/lib/python3.9/site-packages/cloudpickle/cloudpickle_fast.py:602, in CloudPickler.dump(self, obj)
600 def dump(self, obj):
601 try:
--> 602 return Pickler.dump(self, obj)
603 except RuntimeError as e:
604 if "recursion" in e.args[0]:
File ~/work/larch/larch/larch/larch/numba/model.py:1927, in NumbaModel.__getstate__(self)
1921 def __getstate__(self):
1922 state = dict(
1923 float_dtype=self.float_dtype,
1924 constraint_intensity=self.constraint_intensity,
1925 constraint_sharpness=self.constraint_sharpness,
1926 _constraint_funcs=self._constraint_funcs,
-> 1927 _private__graph=self._private__graph,
1928 )
1929 return super().__getstate__(), state
AttributeError: 'NumbaModel' object has no attribute '_private__graph'