def _convert_to_long_format(self):
"""
Convert data to the long format expected by pylogit
"""
model = self.model
(alt_specific_vars,
availability_vars) = self._encode_alternatives_as_integers()
# Create observation number column as a range of integers from 1
model.data[_OBSERVATION_COL] = np.arange(model.data.shape[0],
dtype=int)+1
# Use pylogit routine to convert to long format
self.long_data = pl.convert_wide_to_long(
wide_data=model.data,
ind_vars=model.alternative_independent_variables,
alt_specific_vars=alt_specific_vars,
availability_vars=availability_vars,
obs_id_col=_OBSERVATION_COL,
choice_col=_CHOICE_COL,
new_alt_id_name=_CHOICE_ID_COL
)
# Remove choice bool and observation_id column from wide data
model.data.drop(columns=[_CHOICE_COL, _OBSERVATION_COL],
inplace=True)
def convert_wide_to_long():
# Print a status message
print("Beginning Conversion Process.")
# Load the wide format car data
wide_car = pd.read_csv("../../data/raw/car_wide_format.csv")
# Create the list of individual specific variables
ind_variables = wide_car.columns.tolist()[1:4]
# Specify the variables that vary across individuals and some or all
# alternatives.
# The keys are the column names that will be used in the long format df.
# The values are dictionaries whose key-value pairs are the alternative id
# and the column name of the corresponding column that encodes that
# variable for the given alternative. Examples below.
new_name_to_old_base = {
'body_type': 'type{}',
'fuel_type': 'fuel{}',
'price_over_log_income': 'price{}',
'range': 'range{}',
'acceleration': 'acc{}',
'top_speed': 'speed{}',
'pollution': 'pollution{}',
'vehicle_size': 'size{}',
'luggage_space': 'space{}',
'cents_per_mile': 'cost{}',
'station_availability': 'station{}'
}
alt_varying_variables =\
{k: dict([(x, v.format(x)) for x in range(1, 7)])
for k, v in list(new_name_to_old_base.items())}
# Specify the availability variables
# Note that the keys of the dictionary are the alternative id's.
# The values are the columns denoting the availability for the
# given mode in the dataset.
availability_variables =\
{x: 'avail_{}'.format(x) for x in range(1, 7)}
for col in availability_variables.values():
wide_car[col] = 1
##########
# Determine the columns for: alternative ids, the observation ids, and the choice
##########
# The 'custom_alt_id' is the name of a column to be created in the long-
# format data. It will identify the alternative associated with each row.
custom_alt_id = "alt_id"
# Create a custom id column that ignores the fact that this is a
# panel/repeated-observations dataset. Note the +1 ensures the id's start
# at one.
obs_id_column = "obs_id"
wide_car[obs_id_column] = np.arange(1, wide_car.shape[0] + 1, dtype=int)
# Create a variable recording the choice column
choice_column = "choice"
# Store the original choice column in a new variable
wide_car['orig_choices'] = wide_car['choice'].values
# Alter the original choice column
choice_str_to_value = {'choice{}'.format(x): x for x in range(1, 7)}
wide_car[choice_column] = wide_car[choice_column].map(choice_str_to_value)
# Convert the wide-format data to long format
long_car =\
pl.convert_wide_to_long(wide_data=wide_car,
ind_vars=ind_variables,
alt_specific_vars=alt_varying_variables,
availability_vars=availability_variables,
obs_id_col=obs_id_column,
choice_col=choice_column,
new_alt_id_name=custom_alt_id)
# Save the long-format data to file.
long_car.to_csv("../../data/interim/car_long_format.csv", index=False)
# Print a status message.
print("Finished Conversion Process.")
return None
# Simulate data without parents
sim_bike_data_no_parent = sim.sim_node_no_parent(
params_dict=bike_data_params, size=sim_size)
sim_bike_data_wide = copy.deepcopy(sim_bike_data_no_parent)
# Simulate nodes based on causal graphs
exec(SIMULATE_NODES_WIDE)
# Converting Data from Wide to Long
print("Converting data from wide to long...")
long_sim_data = pl.convert_wide_to_long(
wide_data=sim_bike_data_wide,
ind_vars=IND_VARIABLES,
alt_specific_vars=ALT_VARYING_VARIABLES,
availability_vars=AVAILABILITY_VARIABLES,
obs_id_col=OBS_ID_COL,
choice_col=CHOICE_COL,
new_alt_id_name=CUSTOM_ALT_ID,
)
# Create a cars per licensed drivers column
long_sim_data["cars_per_licensed_drivers"] = 0
long_sim_data.loc[
long_sim_data.num_licensed_drivers > 0,
"cars_per_licensed_drivers"] = long_sim_data.num_cars / long_sim_data.num_licensed_drivers.astype(
float)
# Add a variable representing cost divided by distance
long_sim_data["cost_per_distance"] = 0
long_sim_data.loc[
def convert_wide_to_long(*args, **kwargs):
return pylogit.convert_wide_to_long(*args, **kwargs)
fishing_ind_variables = ["site_choice", "boat", "kids", "ones"]
fishing_availability_variables = OrderedDict()
for alt_id, var in zip(range(1, 101),
["availability_{}".format(i) for i in range(1, 101)]):
fishing_availability_variables[alt_id] = var
# In[105]:
##########
# Actually perform the conversion to long format
##########
long_fishing_df = pl.convert_wide_to_long(
wide_data=data,
ind_vars=fishing_ind_variables,
alt_specific_vars=fishing_alt_varying_variables,
availability_vars=fishing_availability_variables,
obs_id_col=fishing_obs_id_col,
choice_col=fishing_choice_column,
new_alt_id_name=fishing_alt_id)
# 240400 rows × 14 columns
# Look at the long format Heating data
long_fishing_df.head()
long_fishing_df.to_csv("long_data.csv")
# In[84]:
long_fishing_df
# In[ ]:
# In[ ]:
alt_varying_variables = {
u'walk_time':
dict([(2, 'tt_walk'), (3, 'walkTime_PT')]),
u'vehicle_time':
dict([(0, 'tt_drive'), (3, 'transitTime_PT')]),
u'cycle_time':
dict([(1, 'tt_cycle')]),
u'cost':
dict([(0, 'cost_drive'), (1, 'cost_cycle'), (2, 'cost_walk'),
(3, 'cost_PT')])
}
longSimPop = pl.convert_wide_to_long(simPop,
ind_variables,
alt_varying_variables,
availability_variables,
obs_id_column,
choice_column,
new_alt_id_name=custom_alt_id)
##################### Create the model specification #####################
basic_specification = OrderedDict()
basic_names = OrderedDict()
basic_specification["intercept"] = [1, 2, 3]
basic_names["intercept"] = ['ASC Walk', 'ASC Cycle', 'ASC Transit']
#basic_specification["lwBalance_home"] = [1, 2, 3]
#basic_names["lwBalance_home"] = [
# 'lwBalance_home (Cycle)',
u'mpg':
dict([(1, 'EV_mpg'), (2, 'PHEV_mpg'), (3, 'CONV_mpg')]),
u'annual_cost':
dict([(1, 'EV_annual_cost'), (2, 'PHEV_annual_cost'),
(3, 'CONV_annual_cost')]),
u'parking_avail':
dict([(1, 'EV_parking_avail'), (2, 'PHEV_parking_avail'),
(3, 'CONV_parking_avail')]),
u'charging_hrs':
dict([(1, 'EV_charging_hrs'), (2, 'PHEV_charging_hrs'),
(3, 'CONV_charging_hrs')])
}
availability_variables = {1: 'EV_AV', 2: 'PHEV_AV', 3: 'CONV_AV'}
custom_alt_id = "mode_id"
obs_id_column = "custom_id"
dataWide[obs_id_column] = np.arange(dataWide.shape[0], dtype=int) + 1
choice_column = "Choice"
dataWide[choice_column] = dataWide[choice_column].astype('int')
dataLong = pl.convert_wide_to_long(dataWide,
ind_variables,
alt_varying_variables,
availability_variables,
obs_id_column,
choice_column,
new_alt_id_name=custom_alt_id)
# TODO: ADD YOUR MODEL ESTIMATION SCRIPT HERE
def missingToUsable(missing, len_):
return_arr = np.ones(len_)
return_arr[list(missing)] = 0
return_idx = np.where(return_arr == 1)[0]
return (return_idx)
missing = missingIndexes()
usable = missingToUsable(missing, len(raw_data))
long_data = pl.convert_wide_to_long(raw_data,
columns,
alt_varying_variables,
availability_variables,
obs_id_column,
choice_column,
new_alt_id_name=custom_alt_id)
missing_long = missingIndexes(stack=long_data)
usable_long = missingToUsable(missing_long, len(long_data))
def addVariable(var_name,
modes,
desc_array,
names=names,
specification=specification):
specification[var_name] = modes
names[var_name] = desc_array
def unpack_data_for_pylogit(self, df, entnames):
"""
Unpacks a Multi-indexed pandas DataFrame in a way that can be consumed
by pylogit in long-format.
Inputs:
-------
df: DataFrame
DataFrame with multi-index where each index is an entity and object
of comparison made, for this code to be triggered these would also
contain some features.
entnames: list of strings
The name of the columns that contain the entity names.
Returns:
--------
long_x_comb: DataFrame
The DataFrame ready for pylogit modelling in long-format
"""
availability_vars = {}
# Get the distinct entities
all_unique_indices = get_distinct_entities(df)
x_comb = df.reset_index()
ind_variables = df.drop([self.target_col_name],
axis=1).columns.tolist()
for name in entnames:
x_comb[name] = x_comb[name].map(self.rplc_lkp)
for i in all_unique_indices:
colname = str(i) + '_AV'
x_comb[colname] = np.where(
(x_comb[entnames[0]] == self.rplc_lkp[i]) |
(x_comb[entnames[1]] == self.rplc_lkp[i]), 1, 0)
availability_vars[self.rplc_lkp[i]] = colname
x_comb['CHOICE'] = np.where(x_comb[self.target_col_name] == 1,
x_comb[entnames[0]], x_comb[entnames[1]])
x_comb.drop(self.target_col_name, axis=1, inplace=True)
custom_alt_id = 'entity'
obs_id_column = 'observation'
x_comb[obs_id_column] = np.arange(x_comb.shape[0], dtype=int) + 1
choice_column = "CHOICE"
x_comb.drop(entnames, axis=1, inplace=True)
long_x_comb = pl.convert_wide_to_long(x_comb,
ind_variables, {},
availability_vars,
obs_id_column,
choice_column,
new_alt_id_name=custom_alt_id)
return long_x_comb
6: 'K_AV',
7: 'W_AV'
}
alt_id_col = 'ALT_ID'
# 'ID' is the decision-maker, but there are multiple observations (choice scenarios) for each
df['OBS_ID'] = np.arange(df.shape[0], dtype=int) + 1
obs_id_col = 'OBS_ID'
choice_col = 'CHOICE'
data = pylogit.convert_wide_to_long(df,
ind_vars,
alt_varying_vars,
availability_vars,
obs_id_col,
choice_col,
new_alt_id_name=alt_id_col)
# Specify important columns
ind_id_col = 'ID'
obs_id_col = 'OBS_ID'
alt_id_col = 'ALT_ID'
choice_col = 'CHOICE'
# base case is alt_id = 7 (walking)
spec = OrderedDict([('intercept', [1, 2, 3, 4, 5, 6]),
('travel_time', [[1, 2], [3, 4, 5], [6, 7]]),
def generate_data(train=True):
wide_lpmc = pd.read_table("lpmc.dat", sep="\t")
#%%
# define travel time for public transport
#########################################
wide_lpmc["dur_pt"] = wide_lpmc["dur_pt_access"] + wide_lpmc[
"dur_pt_rail"] + wide_lpmc["dur_pt_bus"] + wide_lpmc["dur_pt_int"]
# define cost for driving alternative
#####################################
wide_lpmc['cost_driving'] = wide_lpmc['cost_driving_fuel'] + wide_lpmc[
'cost_driving_ccharge']
# define avaiblity columns
##########################
wide_lpmc['AV'] = np.ones(len(wide_lpmc['cost_driving']))
###############################################
# SEGMENTATION WRT GENDER, PURPOSE, AGE, SEASON
###############################################
'''
we copy the columns into segmented ones and put 0
when the gender does not match the gender considered in the column.
'''
# Driving cost male:
wide_lpmc['cost_driving_male_age1_cold_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_cold_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_cold_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_cold_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_cold_purpose5'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_warm_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_warm_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_warm_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_warm_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age1_warm_purpose5'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_cold_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_cold_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_cold_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_cold_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_cold_purpose5'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_warm_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_warm_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_warm_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_warm_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age2_warm_purpose5'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_cold_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_cold_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_cold_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_cold_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_cold_purpose5'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_warm_purpose1'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_warm_purpose2'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_warm_purpose3'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_warm_purpose4'] = wide_lpmc[
'cost_driving']
wide_lpmc['cost_driving_male_age3_warm_purpose5'] = wide_lpmc[
'cost_driving']
# Public transport cost male:
wide_lpmc['cost_pt_male_age1_cold_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_cold_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_cold_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_cold_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_cold_purpose5'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_warm_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_warm_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_warm_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_warm_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age1_warm_purpose5'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_cold_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_cold_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_cold_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_cold_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_cold_purpose5'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_warm_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_warm_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_warm_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_warm_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age2_warm_purpose5'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_cold_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_cold_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_cold_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_cold_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_cold_purpose5'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_warm_purpose1'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_warm_purpose2'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_warm_purpose3'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_warm_purpose4'] = wide_lpmc['cost_transit']
wide_lpmc['cost_pt_male_age3_warm_purpose5'] = wide_lpmc['cost_transit']
# Driving time male:
wide_lpmc['dur_driving_male_age1_cold_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_cold_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_cold_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_cold_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_cold_purpose5'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_warm_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_warm_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_warm_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_warm_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age1_warm_purpose5'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_cold_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_cold_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_cold_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_cold_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_cold_purpose5'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_warm_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_warm_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_warm_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_warm_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age2_warm_purpose5'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_cold_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_cold_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_cold_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_cold_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_cold_purpose5'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_warm_purpose1'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_warm_purpose2'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_warm_purpose3'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_warm_purpose4'] = wide_lpmc['dur_driving']
wide_lpmc['dur_driving_male_age3_warm_purpose5'] = wide_lpmc['dur_driving']
# Public transport time male:
wide_lpmc['dur_pt_male_age1_cold_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_cold_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_cold_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_cold_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_cold_purpose5'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_warm_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_warm_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_warm_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_warm_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age1_warm_purpose5'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_cold_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_cold_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_cold_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_cold_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_cold_purpose5'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_warm_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_warm_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_warm_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_warm_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age2_warm_purpose5'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_cold_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_cold_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_cold_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_cold_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_cold_purpose5'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_warm_purpose1'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_warm_purpose2'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_warm_purpose3'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_warm_purpose4'] = wide_lpmc['dur_pt']
wide_lpmc['dur_pt_male_age3_warm_purpose5'] = wide_lpmc['dur_pt']
# Walking time male:
wide_lpmc['dur_walking_male_age1_cold_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_cold_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_cold_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_cold_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_cold_purpose5'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_warm_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_warm_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_warm_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_warm_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age1_warm_purpose5'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_cold_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_cold_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_cold_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_cold_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_cold_purpose5'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_warm_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_warm_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_warm_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_warm_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age2_warm_purpose5'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_cold_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_cold_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_cold_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_cold_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_cold_purpose5'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_warm_purpose1'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_warm_purpose2'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_warm_purpose3'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_warm_purpose4'] = wide_lpmc['dur_walking']
wide_lpmc['dur_walking_male_age3_warm_purpose5'] = wide_lpmc['dur_walking']
# Cycling time male:
wide_lpmc['dur_cycling_male_age1_cold_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_cold_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_cold_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_cold_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_cold_purpose5'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_warm_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_warm_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_warm_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_warm_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age1_warm_purpose5'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_cold_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_cold_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_cold_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_cold_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_cold_purpose5'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_warm_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_warm_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_warm_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_warm_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age2_warm_purpose5'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_cold_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_cold_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_cold_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_cold_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_cold_purpose5'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_warm_purpose1'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_warm_purpose2'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_warm_purpose3'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_warm_purpose4'] = wide_lpmc['dur_cycling']
wide_lpmc['dur_cycling_male_age3_warm_purpose5'] = wide_lpmc['dur_cycling']
# Distance male:
wide_lpmc['distance_male_age1_cold_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_cold_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_cold_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_cold_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_cold_purpose5'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_warm_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_warm_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_warm_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_warm_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age1_warm_purpose5'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_cold_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_cold_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_cold_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_cold_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_cold_purpose5'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_warm_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_warm_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_warm_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_warm_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age2_warm_purpose5'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_cold_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_cold_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_cold_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_cold_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_cold_purpose5'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_warm_purpose1'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_warm_purpose2'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_warm_purpose3'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_warm_purpose4'] = wide_lpmc['distance']
wide_lpmc['distance_male_age3_warm_purpose5'] = wide_lpmc['distance']
###########
## Def Lambdas
######
# Distance male:
wide_lpmc['lambda_male_age1_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age1_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age2_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_male_age3_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age1_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age2_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_time_male_age3_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age1_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age2_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_cold_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_cold_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_cold_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_cold_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_cold_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_warm_purpose1'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_warm_purpose2'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_warm_purpose3'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_warm_purpose4'] = np.ones(
len(wide_lpmc['cost_driving']))
wide_lpmc['lambda_cost_male_age3_warm_purpose5'] = np.ones(
len(wide_lpmc['cost_driving']))
#%% FILTERS
filter_age1 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("age1")]
filter_age2 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("age2")]
filter_age3 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("age3")]
wide_lpmc.loc[(wide_lpmc['age'] > 18), filter_age1] = 0
wide_lpmc.loc[((wide_lpmc['age'] <= 18) & (wide_lpmc['age'] > 64)),
filter_age2] = 0
wide_lpmc.loc[(wide_lpmc['age'] <= 64), filter_age3] = 0
filter_warm = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("warm")]
filter_cold = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("cold")]
wide_lpmc.loc[((wide_lpmc['travel_month'] != 12) &
(wide_lpmc['travel_month'] > 2)), filter_cold] = 0
wide_lpmc.loc[((wide_lpmc['travel_month'] == 12) |
(wide_lpmc['travel_month'] == 1) |
(wide_lpmc['travel_month'] == 2)), filter_warm] = 0
filter_p1 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("purpose1")]
filter_p2 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("purpose2")]
filter_p3 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("purpose3")]
filter_p4 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("purpose4")]
filter_p5 = wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("purpose5")]
wide_lpmc.loc[(wide_lpmc['purpose'] != 1), filter_p1] = 0
wide_lpmc.loc[(wide_lpmc['purpose'] != 2), filter_p2] = 0
wide_lpmc.loc[(wide_lpmc['purpose'] != 3), filter_p3] = 0
wide_lpmc.loc[(wide_lpmc['purpose'] != 4), filter_p4] = 0
wide_lpmc.loc[(wide_lpmc['purpose'] != 5), filter_p5] = 0
#%%
# delete useless columns (can be discussed)
##############################################
del wide_lpmc['dur_pt_access']
del wide_lpmc['dur_pt_rail']
del wide_lpmc['dur_pt_bus']
del wide_lpmc['dur_pt_int']
del wide_lpmc['cost_driving_fuel']
del wide_lpmc['cost_driving_ccharge']
del wide_lpmc['driving_license']
del wide_lpmc['car_ownership']
del wide_lpmc['trip_id']
del wide_lpmc['household_id']
del wide_lpmc['person_n']
del wide_lpmc['trip_n']
del wide_lpmc['purpose']
del wide_lpmc['fueltype']
del wide_lpmc['faretype']
del wide_lpmc['bus_scale']
del wide_lpmc['female']
del wide_lpmc['pt_interchanges']
del wide_lpmc['driving_traffic_percent']
if (train):
wide_lpmc = wide_lpmc.drop(
wide_lpmc[wide_lpmc['survey_year'] != 3].index)
else:
wide_lpmc = wide_lpmc.drop(
wide_lpmc[wide_lpmc['survey_year'] == 3].index)
del wide_lpmc['survey_year']
del wide_lpmc['travel_year']
del wide_lpmc['travel_month']
del wide_lpmc['travel_date']
del wide_lpmc['day_of_week']
del wide_lpmc['start_time']
del wide_lpmc['dur_driving']
del wide_lpmc['dur_walking']
del wide_lpmc['dur_pt']
del wide_lpmc['dur_cycling']
del wide_lpmc['distance']
del wide_lpmc['cost_driving']
del wide_lpmc['cost_transit']
del wide_lpmc['age']
#%%
''' ind_vars : list of strings.
Each element should be a column heading in `wide_data` that denotes a
variable that varies across observations but not across alternatives.
'''
ind_variables = list(wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("distance")])
ind_variables = ind_variables + list(wide_lpmc.columns[pd.Series(
wide_lpmc.columns).str.contains("lambda")])
#ind_variables=ind_variables+list(wide_lpmc['survey_year'])
#%%
alt_varying_variables = {
u'travel_time_male_age1_cold_purpose1':
dict([(1, 'dur_walking_male_age1_cold_purpose1'),
(2, 'dur_cycling_male_age1_cold_purpose1'),
(3, 'dur_pt_male_age1_cold_purpose1'),
(4, 'dur_driving_male_age1_cold_purpose1')]),
u'travel_time_male_age1_cold_purpose2':
dict([(1, 'dur_walking_male_age1_cold_purpose2'),
(2, 'dur_cycling_male_age1_cold_purpose2'),
(3, 'dur_pt_male_age1_cold_purpose2'),
(4, 'dur_driving_male_age1_cold_purpose2')]),
u'travel_time_male_age1_cold_purpose3':
dict([(1, 'dur_walking_male_age1_cold_purpose3'),
(2, 'dur_cycling_male_age1_cold_purpose3'),
(3, 'dur_pt_male_age1_cold_purpose3'),
(4, 'dur_driving_male_age1_cold_purpose3')]),
u'travel_time_male_age1_cold_purpose4':
dict([(1, 'dur_walking_male_age1_cold_purpose4'),
(2, 'dur_cycling_male_age1_cold_purpose4'),
(3, 'dur_pt_male_age1_cold_purpose4'),
(4, 'dur_driving_male_age1_cold_purpose4')]),
u'travel_time_male_age1_cold_purpose5':
dict([(1, 'dur_walking_male_age1_cold_purpose5'),
(2, 'dur_cycling_male_age1_cold_purpose5'),
(3, 'dur_pt_male_age1_cold_purpose5'),
(4, 'dur_driving_male_age1_cold_purpose5')]),
u'travel_time_male_age1_warm_purpose1':
dict([(1, 'dur_walking_male_age1_warm_purpose1'),
(2, 'dur_cycling_male_age1_warm_purpose1'),
(3, 'dur_pt_male_age1_warm_purpose1'),
(4, 'dur_driving_male_age1_warm_purpose1')]),
u'travel_time_male_age1_warm_purpose2':
dict([(1, 'dur_walking_male_age1_warm_purpose2'),
(2, 'dur_cycling_male_age1_warm_purpose2'),
(3, 'dur_pt_male_age1_warm_purpose2'),
(4, 'dur_driving_male_age1_warm_purpose2')]),
u'travel_time_male_age1_warm_purpose3':
dict([(1, 'dur_walking_male_age1_warm_purpose3'),
(2, 'dur_cycling_male_age1_warm_purpose3'),
(3, 'dur_pt_male_age1_warm_purpose3'),
(4, 'dur_driving_male_age1_warm_purpose3')]),
u'travel_time_male_age1_warm_purpose4':
dict([(1, 'dur_walking_male_age1_warm_purpose4'),
(2, 'dur_cycling_male_age1_warm_purpose4'),
(3, 'dur_pt_male_age1_warm_purpose4'),
(4, 'dur_driving_male_age1_warm_purpose4')]),
u'travel_time_male_age1_warm_purpose5':
dict([(1, 'dur_walking_male_age1_warm_purpose5'),
(2, 'dur_cycling_male_age1_warm_purpose5'),
(3, 'dur_pt_male_age1_warm_purpose5'),
(4, 'dur_driving_male_age1_warm_purpose5')]),
u'travel_time_male_age2_cold_purpose1':
dict([(1, 'dur_walking_male_age2_cold_purpose1'),
(2, 'dur_cycling_male_age2_cold_purpose1'),
(3, 'dur_pt_male_age2_cold_purpose1'),
(4, 'dur_driving_male_age2_cold_purpose1')]),
u'travel_time_male_age2_cold_purpose2':
dict([(1, 'dur_walking_male_age2_cold_purpose2'),
(2, 'dur_cycling_male_age2_cold_purpose2'),
(3, 'dur_pt_male_age2_cold_purpose2'),
(4, 'dur_driving_male_age2_cold_purpose2')]),
u'travel_time_male_age2_cold_purpose3':
dict([(1, 'dur_walking_male_age2_cold_purpose3'),
(2, 'dur_cycling_male_age2_cold_purpose3'),
(3, 'dur_pt_male_age2_cold_purpose3'),
(4, 'dur_driving_male_age2_cold_purpose3')]),
u'travel_time_male_age2_cold_purpose4':
dict([(1, 'dur_walking_male_age2_cold_purpose4'),
(2, 'dur_cycling_male_age2_cold_purpose4'),
(3, 'dur_pt_male_age2_cold_purpose4'),
(4, 'dur_driving_male_age2_cold_purpose4')]),
u'travel_time_male_age2_cold_purpose5':
dict([(1, 'dur_walking_male_age2_cold_purpose5'),
(2, 'dur_cycling_male_age2_cold_purpose5'),
(3, 'dur_pt_male_age2_cold_purpose5'),
(4, 'dur_driving_male_age2_cold_purpose5')]),
u'travel_time_male_age2_warm_purpose1':
dict([(1, 'dur_walking_male_age2_warm_purpose1'),
(2, 'dur_cycling_male_age2_warm_purpose1'),
(3, 'dur_pt_male_age2_warm_purpose1'),
(4, 'dur_driving_male_age2_warm_purpose1')]),
u'travel_time_male_age2_warm_purpose2':
dict([(1, 'dur_walking_male_age2_warm_purpose2'),
(2, 'dur_cycling_male_age2_warm_purpose2'),
(3, 'dur_pt_male_age2_warm_purpose2'),
(4, 'dur_driving_male_age2_warm_purpose2')]),
u'travel_time_male_age2_warm_purpose3':
dict([(1, 'dur_walking_male_age2_warm_purpose3'),
(2, 'dur_cycling_male_age2_warm_purpose3'),
(3, 'dur_pt_male_age2_warm_purpose3'),
(4, 'dur_driving_male_age2_warm_purpose3')]),
u'travel_time_male_age2_warm_purpose4':
dict([(1, 'dur_walking_male_age2_warm_purpose4'),
(2, 'dur_cycling_male_age2_warm_purpose4'),
(3, 'dur_pt_male_age2_warm_purpose4'),
(4, 'dur_driving_male_age2_warm_purpose4')]),
u'travel_time_male_age2_warm_purpose5':
dict([(1, 'dur_walking_male_age2_warm_purpose5'),
(2, 'dur_cycling_male_age2_warm_purpose5'),
(3, 'dur_pt_male_age2_warm_purpose5'),
(4, 'dur_driving_male_age2_warm_purpose5')]),
u'travel_time_male_age3_cold_purpose1':
dict([(1, 'dur_walking_male_age3_cold_purpose1'),
(2, 'dur_cycling_male_age3_cold_purpose1'),
(3, 'dur_pt_male_age3_cold_purpose1'),
(4, 'dur_driving_male_age3_cold_purpose1')]),
u'travel_time_male_age3_cold_purpose2':
dict([(1, 'dur_walking_male_age3_cold_purpose2'),
(2, 'dur_cycling_male_age3_cold_purpose2'),
(3, 'dur_pt_male_age3_cold_purpose2'),
(4, 'dur_driving_male_age3_cold_purpose2')]),
u'travel_time_male_age3_cold_purpose3':
dict([(1, 'dur_walking_male_age3_cold_purpose3'),
(2, 'dur_cycling_male_age3_cold_purpose3'),
(3, 'dur_pt_male_age3_cold_purpose3'),
(4, 'dur_driving_male_age3_cold_purpose3')]),
u'travel_time_male_age3_cold_purpose4':
dict([(1, 'dur_walking_male_age3_cold_purpose4'),
(2, 'dur_cycling_male_age3_cold_purpose4'),
(3, 'dur_pt_male_age3_cold_purpose4'),
(4, 'dur_driving_male_age3_cold_purpose4')]),
u'travel_time_male_age3_cold_purpose5':
dict([(1, 'dur_walking_male_age3_cold_purpose5'),
(2, 'dur_cycling_male_age3_cold_purpose5'),
(3, 'dur_pt_male_age3_cold_purpose5'),
(4, 'dur_driving_male_age3_cold_purpose5')]),
u'travel_time_male_age3_warm_purpose1':
dict([(1, 'dur_walking_male_age3_warm_purpose1'),
(2, 'dur_cycling_male_age3_warm_purpose1'),
(3, 'dur_pt_male_age3_warm_purpose1'),
(4, 'dur_driving_male_age3_warm_purpose1')]),
u'travel_time_male_age3_warm_purpose2':
dict([(1, 'dur_walking_male_age3_warm_purpose2'),
(2, 'dur_cycling_male_age3_warm_purpose2'),
(3, 'dur_pt_male_age3_warm_purpose2'),
(4, 'dur_driving_male_age3_warm_purpose2')]),
u'travel_time_male_age3_warm_purpose3':
dict([(1, 'dur_walking_male_age3_warm_purpose3'),
(2, 'dur_cycling_male_age3_warm_purpose3'),
(3, 'dur_pt_male_age3_warm_purpose3'),
(4, 'dur_driving_male_age3_warm_purpose3')]),
u'travel_time_male_age3_warm_purpose4':
dict([(1, 'dur_walking_male_age3_warm_purpose4'),
(2, 'dur_cycling_male_age3_warm_purpose4'),
(3, 'dur_pt_male_age3_warm_purpose4'),
(4, 'dur_driving_male_age3_warm_purpose4')]),
u'travel_time_male_age3_warm_purpose5':
dict([(1, 'dur_walking_male_age3_warm_purpose5'),
(2, 'dur_cycling_male_age3_warm_purpose5'),
(3, 'dur_pt_male_age3_warm_purpose5'),
(4, 'dur_driving_male_age3_warm_purpose5')]),
u'travel_cost_male_age1_cold_purpose1':
dict([(3, 'cost_pt_male_age1_cold_purpose1'),
(4, 'cost_driving_male_age1_cold_purpose1')]),
u'travel_cost_male_age1_cold_purpose2':
dict([(3, 'cost_pt_male_age1_cold_purpose2'),
(4, 'cost_driving_male_age1_cold_purpose2')]),
u'travel_cost_male_age1_cold_purpose3':
dict([(3, 'cost_pt_male_age1_cold_purpose3'),
(4, 'cost_driving_male_age1_cold_purpose3')]),
u'travel_cost_male_age1_cold_purpose4':
dict([(3, 'cost_pt_male_age1_cold_purpose4'),
(4, 'cost_driving_male_age1_cold_purpose4')]),
u'travel_cost_male_age1_cold_purpose5':
dict([(3, 'cost_pt_male_age1_cold_purpose5'),
(4, 'cost_driving_male_age1_cold_purpose5')]),
u'travel_cost_male_age1_warm_purpose1':
dict([(3, 'cost_pt_male_age1_warm_purpose1'),
(4, 'cost_driving_male_age1_warm_purpose1')]),
u'travel_cost_male_age1_warm_purpose2':
dict([(3, 'cost_pt_male_age1_warm_purpose2'),
(4, 'cost_driving_male_age1_warm_purpose2')]),
u'travel_cost_male_age1_warm_purpose3':
dict([(3, 'cost_pt_male_age1_warm_purpose3'),
(4, 'cost_driving_male_age1_warm_purpose3')]),
u'travel_cost_male_age1_warm_purpose4':
dict([(3, 'cost_pt_male_age1_warm_purpose4'),
(4, 'cost_driving_male_age1_warm_purpose4')]),
u'travel_cost_male_age1_warm_purpose5':
dict([(3, 'cost_pt_male_age1_warm_purpose5'),
(4, 'cost_driving_male_age1_warm_purpose5')]),
u'travel_cost_male_age2_cold_purpose1':
dict([(3, 'cost_pt_male_age2_cold_purpose1'),
(4, 'cost_driving_male_age2_cold_purpose1')]),
u'travel_cost_male_age2_cold_purpose2':
dict([(3, 'cost_pt_male_age2_cold_purpose2'),
(4, 'cost_driving_male_age2_cold_purpose2')]),
u'travel_cost_male_age2_cold_purpose3':
dict([(3, 'cost_pt_male_age2_cold_purpose3'),
(4, 'cost_driving_male_age2_cold_purpose3')]),
u'travel_cost_male_age2_cold_purpose4':
dict([(3, 'cost_pt_male_age2_cold_purpose4'),
(4, 'cost_driving_male_age2_cold_purpose4')]),
u'travel_cost_male_age2_cold_purpose5':
dict([(3, 'cost_pt_male_age2_cold_purpose5'),
(4, 'cost_driving_male_age2_cold_purpose5')]),
u'travel_cost_male_age2_warm_purpose1':
dict([(3, 'cost_pt_male_age2_warm_purpose1'),
(4, 'cost_driving_male_age2_warm_purpose1')]),
u'travel_cost_male_age2_warm_purpose2':
dict([(3, 'cost_pt_male_age2_warm_purpose2'),
(4, 'cost_driving_male_age2_warm_purpose2')]),
u'travel_cost_male_age2_warm_purpose3':
dict([(3, 'cost_pt_male_age2_warm_purpose3'),
(4, 'cost_driving_male_age2_warm_purpose3')]),
u'travel_cost_male_age2_warm_purpose4':
dict([(3, 'cost_pt_male_age2_warm_purpose4'),
(4, 'cost_driving_male_age2_warm_purpose4')]),
u'travel_cost_male_age2_warm_purpose5':
dict([(3, 'cost_pt_male_age2_warm_purpose5'),
(4, 'cost_driving_male_age2_warm_purpose5')]),
u'travel_cost_male_age3_cold_purpose1':
dict([(3, 'cost_pt_male_age3_cold_purpose1'),
(4, 'cost_driving_male_age3_cold_purpose1')]),
u'travel_cost_male_age3_cold_purpose2':
dict([(3, 'cost_pt_male_age3_cold_purpose2'),
(4, 'cost_driving_male_age3_cold_purpose2')]),
u'travel_cost_male_age3_cold_purpose3':
dict([(3, 'cost_pt_male_age3_cold_purpose3'),
(4, 'cost_driving_male_age3_cold_purpose3')]),
u'travel_cost_male_age3_cold_purpose4':
dict([(3, 'cost_pt_male_age3_cold_purpose4'),
(4, 'cost_driving_male_age3_cold_purpose4')]),
u'travel_cost_male_age3_cold_purpose5':
dict([(3, 'cost_pt_male_age3_cold_purpose5'),
(4, 'cost_driving_male_age3_cold_purpose5')]),
u'travel_cost_male_age3_warm_purpose1':
dict([(3, 'cost_pt_male_age3_warm_purpose1'),
(4, 'cost_driving_male_age3_warm_purpose1')]),
u'travel_cost_male_age3_warm_purpose2':
dict([(3, 'cost_pt_male_age3_warm_purpose2'),
(4, 'cost_driving_male_age3_warm_purpose2')]),
u'travel_cost_male_age3_warm_purpose3':
dict([(3, 'cost_pt_male_age3_warm_purpose3'),
(4, 'cost_driving_male_age3_warm_purpose3')]),
u'travel_cost_male_age3_warm_purpose4':
dict([(3, 'cost_pt_male_age3_warm_purpose4'),
(4, 'cost_driving_male_age3_warm_purpose4')]),
u'travel_cost_male_age3_warm_purpose5':
dict([(3, 'cost_pt_male_age3_warm_purpose5'),
(4, 'cost_driving_male_age3_warm_purpose5')]),
}
#%%
availability_variables = {1: 'AV', 2: 'AV', 3: 'AV', 4: 'AV'}
custom_alt_id = "mode_id"
# obs_id_column = "trip_id"
# wide_lpmc["trip_id"] =wide_lpmc["trip_id"]+1
obs_id_column = "custom_id"
wide_lpmc[obs_id_column] = np.arange(wide_lpmc.shape[0], dtype=int) + 1
# Create a variable recording the choice column
choice_column = "travel_mode"
long_lpmc = pl.convert_wide_to_long(wide_lpmc,
ind_variables,
alt_varying_variables,
availability_variables,
obs_id_column,
choice_column,
new_alt_id_name=custom_alt_id)
return long_lpmc