def __init__(
self,
config,
df,
labels=['car2go', 'enjoy'],
agg_func='count',
save=False,
norm_per_day=False,
name='ReB_Hour.pdf',
):
fig, ax = plt.subplots(1, 1, figsize=config['figsize'])
for label in labels:
sub_df = df[df.vendor == label]
sub_df_filter = Filter(sub_df, config)
sub_df_split_dict = sub_df_filter.split_WD_WE()
sub_df_we = sub_df_split_dict['WE']
sub_df_wd = sub_df_split_dict['WD']
if agg_func == 'count':
sub_df_we = sub_df_we.groupby('Hour').count()
sub_df_wd = sub_df_wd.groupby('Hour').count()
elif agg_func == 'mean':
sub_df_we = sub_df_we.groupby('Hour').mean()
sub_df_wd = sub_df_wd.groupby('Hour').mean()
else:
print('error')
return
sub_df_we = sub_df_we.div(
self.compute_number_of_days(sub_df_we, norm_per_day))
sub_df_wd = sub_df_wd.div(
self.compute_number_of_days(sub_df_we, norm_per_day))
columns_ref = sub_df_we.columns[0]
sub_df_we = sub_df_we[columns_ref]
sub_df_wd = sub_df_wd[columns_ref]
ax.plot(sub_df_wd.index,
sub_df_wd.values,
color=config['colors'][label],
label='%s Working Days' % label)
ax.plot(sub_df_we.index,
sub_df_we.values,
color=config['colors'][label],
label='%s Week Ends' % label,
linestyle='--')
ax.tick_params(axis='x', rotation=15)
ax.legend()
ax.set_xticks(range(0, 24))
ax.grid()
# ax.set_ylim(0,6000)
ax.set_ylabel('Rentals per hour')
ax.set_xlabel('Hour of day')
if save:
plt.savefig(config['output_plot_path'] + name, bboxinches='tight')
fig.show()
def __init__(
self,
config,
df,
labels=['car2go', 'enjoy'],
agg_func='count',
save=False,
norm_per_day=False,
name='ReB_Hour.pdf',
):
if norm_per_day == False: div = 1
else: div = norm_per_day = len(df.Date.unique())
fig, ax = plt.subplots(1, 1, figsize=config['figsize'])
for label in labels:
sub_df = df[df.vendor == label]
sub_df_filter = Filter(sub_df, config)
sub_df_split_dict = sub_df_filter.split_WD_WE()
sub_df_we = sub_df_split_dict['WE']
sub_df_wd = sub_df_split_dict['WD']
column_ref = df.columns[0]
# c2g_number_of_workingdays = c2g_wd.groupby('Wod').count()[column_ref].sum()
# c2g_number_of_weekends = c2g_we.groupby('Wod').count()[column_ref].sum()
sub_df_we = sub_df_we.groupby('Hour')[column_ref].agg(
agg_func).div(div)
sub_df_wd = sub_df_wd.groupby('Hour')[column_ref].agg(
agg_func).div(div)
ax.plot(sub_df_wd.index,
sub_df_wd.values,
color=config['colors_per_city'][label],
label='%s Working Days' % label)
ax.plot(sub_df_we.index,
sub_df_we.values,
color=config['colors_per_city'][label],
label='%s Week Ends' % label,
linestyle='--')
ax.tick_params(axis='x', rotation=15)
ax.legend()
ax.set_xticks(range(0, 24))
ax.grid()
# ax.set_ylim(0,6000)
ax.set_ylabel('Rentals per day')
ax.set_xlabel('Hour of day')
if save:
plt.savefig(config['output_plot_path'] + 'ReB_Hour.pdf',
bboxinches='tight')
fig.show()
def __init__(self, df, city_map, config, label, save=False):
df_before_filters = df.copy()
filter = Filter(df, config)
filter.remove_fake_bookings_torino()
filter.date_standardization()
df = filter.rentals()
converter = DF2GDF()
df_start = converter.df2gdf_point(df.copy(),
dict(Lon='start_lon',
Lat='start_lat'),
crs=config['crs'])
df_end = converter.df2gdf_point(df.copy(),
dict(Lon='end_lon', Lat='end_lat'),
crs=config['crs'])
df = converter.assign_zone_to_rental(df_start,
df_end,
city_map,
convert_dict_start={
'index_right':
'start_zone_ID',
'zone_name': 'start_zone_name'
},
convert_dict_end={
'index_right': 'end_zone_ID',
'zone_name': 'end_zone_name'
})
df = df[df.start_zone_ID.notna()]
df = df[df.end_zone_ID.notna()]
converter.dump_df(df, label + 'Torino', config['data_path'])
city_map_cleaned = converter.drop_zones_without_bookings(df, city_map)
'''Morning bookings'''
df_m = df[df.Hour.isin(list(range(7, 13)))]
df_m = df_m.groupby('end_zone_ID').count()['_id']
'''Evening'''
df_e = df[df.Hour.isin(list(range(17, 22)))]
df_e = df_e.groupby('end_zone_ID').count()['_id']
city_map_cleaned['actract'] = df_e - df_m
city_map_cleaned = converter.reshape_caselle(city_map_cleaned,
config['caselleID'])
city_map_cleaned = city_map_cleaned.dropna()
fig, ax = plt.subplots()
city_map_cleaned.plot(ax=ax, column='actract', legend=True)
ax.set_xticks([])
ax.set_yticks([])
if save: plt.savefig(config['output_plot_path'] + label + 'Torino.pdf')
fig.show()
def __init__(self, config, df, labels=['car2go', 'enjoy']):
self.df = df
self.config = config
self.datasets = {}
for label in labels:
sub_df = self.df[df.vendor == label]
self.datasets[label] = {}
self.datasets[label]['bookings'] = {}
self.datasets[label]['rentals '] = {}
filter = Filter(sub_df, config)
out_b = filter.split_WD_WE()
self.datasets[label]['bookings'] = out_b
filter.rentals()
out_r = filter.split_WD_WE()
self.datasets[label]['rentals'] = out_r
self.labels = labels
nrows = 120000
config = ReadConfig('config.json')
plt.rcParams.update({'font.size': config['fs']})
col2keep = [
'_id', 'city', 'distance', 'driving_distance', 'driving_duration',
'duration', 'end_lat', 'end_lon', 'final_address', 'final_date',
'final_fuel', 'final_time', 'init_address', 'init_date', 'init_fuel',
'init_time', 'plate', 'pt_arrival_date', 'pt_arrival_time',
'pt_distance', 'pt_duration', 'start_lat', 'start_lon', 'vendor',
'walking_distance', 'walking_duration'
]
c2g = Loader(config, 'Torino', 'car2go', nrows)
c2g = c2g[col2keep]
c2g_filter = Filter(c2g, config)
c2g_filter.date_standardization(fmt='%Y-%m-%d %H:%M:%S')
c2g_filter.remove_fake_bookings_torino()
c2g_allevents = c2g.copy()
c2g_filter.rentals()
c2g = c2g_filter.df
enj = Loader(config, 'Torino', 'enjoy', nrows)
enj = enj[col2keep]
enj_filter = Filter(enj, config)
enj_filter.date_standardization(fmt='%Y-%m-%d %H:%M:%S')
enj_filter.remove_fake_bookings_torino()
enj_allevents = enj.copy()
enj_filter.rentals()
enj = enj_filter.df
minn = minn.rename(columns=new_columns)
not_formatted_init = minn[minn.init_date.str.contains('/')].index
minn.loc[not_formatted_init, 'init_date'] = minn.init_date.str.replace(
'/', '-').str.replace(' ', 'T')
minn.loc[not_formatted_init,
'init_date'] = minn.loc[not_formatted_init, 'init_date'] + '.000Z'
not_formatted_final = minn[minn.final_date.str.contains('/')].index
minn.loc[not_formatted_init, 'final_date'] = minn.init_date.str.replace(
'/', '-').str.replace(' ', 'T')
minn.loc[not_formatted_init,
'final_date'] = minn.loc[not_formatted_init,
'final_date'] + '.000Z'
minn_filter = Filter(minn, config)
minn_filter.date_standardization(fmt='%Y-%m-%dT%H:%M:%S.000Z')
minn_filter.localize_timezone('Date_index', 'UTC', 'America/Chicago')
minn_filter.limit_date()
minn_allevents = minn.copy()
minn = minn_filter.df
minn['Date'] = minn.Date_index.dt.date
minn_allevents['Date'] = minn_allevents.Date_index.dt.date
louis = Loader(config, l2, 'car2go', nrows)
louis = louis[~louis.StartTime.str.contains('24:')]
louis['StartTime'] = louis['StartDate'] + 'T' + louis[
'StartTime'] + ':00.000Z'
louis['vendor'] = 'Louisville'
louis['TripDuration'] = louis['TripDuration'].mul(60)