def plot_cluster_track(self, bbox):
"""
Plots graph for users in selected area
:param bbox: coordinates of top-left and bottom-right angles of area
:type bbox: List[List[float]]
:return: None
"""
data = self._get_data_from_plot(bbox)
data_agg = get_all_agg(data, ['trans_count'])
plot.plot_graph(data_agg, 'trans_count', {'export_folder': self.export_folder})
def export_tracks(df, settings, users='all', task='lost', order='all', treshold=0.5,
start_event=None, end_event=None):
settings = check_folder(settings)
export_folder = settings['export_folder']
if task == 'lost' and start_event is None:
settings['start_event'] = 'welcome_see_screen'
# else:
# settings['start_event'] = 'start'
# df = df.sort_values(['user_pseudo_id', 'event_timestamp'])
# first = df.groupby('user_pseudo_id', as_index=False).first()
# first.event_timestamp -= 1
# first.event_name = 'start'
# df = df.append(first, ignore_index=True, sort=False)
agg_list = ['trans_count', 'dt_mean', 'dt_median', 'dt_min', 'dt_max']
if type(users) != str:
df = df[df.user_pseudo_id.isin(users)]
settings['users']['userlist'] = list(users)
else:
if settings.get('users') is None:
settings['users'] = {}
settings['users']['userlist'] = 'all'
df = get_session(df, order=order, treshold=treshold)
if settings.get('events') is None:
settings['events'] = {}
settings['events']['session_order'] = order
settings['total_count'] = df.user_pseudo_id.nunique()
df = get_all_agg(df, agg_list)
settings['events']['session_thr_time'] = treshold * 1e8 * 36
for i in os.listdir(settings['export_folder']):
if 'settings' in i:
set_name = i
with open(os.path.join(settings['export_folder'], set_name), 'w') as f:
json.dump(settings, f)
graph_name = 'graph_{}.csv'.format(datetime.now())
df.to_csv(os.path.join(export_folder, graph_name), index=False)
return export_folder, graph_name, set_name
def export_tracks(df, settings, users='all', task='lost', order='all', treshold=0.5,
start_event=None, end_event=None):
"""
Visualize trajectories from event clickstream (with Mathematica)
:param df: event clickstream
:param settings: experiment config (can be empty dict here)
:param users: `all` or list of user ids to plot specific group
:param task: type of task for different visualization (can be `lost` or `prunned_welcome`)
:param order: depth in sessions for filtering
:param threshold: threshold for session splitting
:param start_event: name of start event in trajectory
:param end_event: name of last event in trajectory
:param df: pd.DataFrame
:param settings: dict
:param users: str or list
:param task: str
:param order: int
:param threshold: float
:param start_event: str
:param end_event: str
:return: None
"""
settings = _check_folder(settings)
export_folder = settings['export_folder']
if task == 'lost' and start_event is None:
settings['start_event'] = 'welcome_see_screen'
# else:
# settings['start_event'] = 'start'
# df = df.sort_values(['user_pseudo_id', 'event_timestamp'])
# first = df.groupby('user_pseudo_id', as_index=False).first()
# first.event_timestamp -= 1
# first.event_name = 'start'
# df = df.append(first, ignore_index=True, sort=False)
agg_list = ['trans_count', 'dt_mean', 'dt_median', 'dt_min', 'dt_max']
if type(users) != str:
df = df[df.user_pseudo_id.isin(users)]
settings['users']['userlist'] = list(users)
else:
if settings.get('users') is None:
settings['users'] = {}
settings['users']['userlist'] = 'all'
df = _get_session(df, order=order, treshold=treshold)
if settings.get('events') is None:
settings['events'] = {}
settings['events']['session_order'] = order
settings['total_count'] = df.user_pseudo_id.nunique()
df = get_all_agg(df, agg_list)
settings['events']['session_thr_time'] = treshold * 1e8 * 36
for i in os.listdir(settings['export_folder']):
if 'settings' in i:
set_name = i
with open(os.path.join(settings['export_folder'], set_name), 'w') as f:
json.dump(settings, f)
graph_name = 'graph_{}.csv'.format(datetime.now())
df.to_csv(os.path.join(export_folder, graph_name), index=False)
return export_folder, graph_name, set_name
def plot_cluster_track(self, bbox):
data = self._get_data_from_plot(bbox)
data_agg = get_all_agg(data, ['trans_count'])
plot_graph_python(data_agg, 'trans_count',
{'export_folder': self.export_folder})