Example 03 - Introduction to SPaT Data#
This example will introduce:
Load the parsed Signal Phase and Timing data with
mtldp.utils.data_ioBasic classes for the SPaT data
How to get SPaT information for a specific time period over a range of days at the corridor, intersection, and movement level
After the SPaT data is parsed from original CSV files, you can load the them using the data platform (mtldp) package. The following flow chart illustrates how the data platform loads and manages the SPaT data.
[3]:
from mtldp.utils.common import show_attributes_and_methods
Load SPaT data#
Load region SPaT history from pickle file#
[4]:
from mtldp.utils.data_io import load_region_spat_from_pickle
region_spat = load_region_spat_from_pickle('data/spat.pickle')
[5]:
show_attributes_and_methods(region_spat)
attributes: {'version_history_df', 'ring_structure_df', 'region_id', 'timezone', 'tod_event_df', 'network_mapping_df', 'node_id_ls', 'node_controller_dict', 'controller_spat_history_dict', 'dial_split_df', 'program_week_df', 'program_day_df'}
methods: {'get_corridor_spat_period', 'get_controller_spat_period', 'get_controller_spat_history', 'get_tod_event_time_splits'}
Get SPaT history of one controller from region#
[6]:
node_id = region_spat.node_id_ls[0]
controller_spat = region_spat.get_controller_spat_history(node_id=node_id)
[7]:
show_attributes_and_methods(controller_spat)
attributes: {'version_history', 'version_history_df', 'ring_structure_df', 'phase_mappings', 'timezone', 'movement_mappings', 'tod_event_df', 'dial_split_versions', 'node_id_ls', 'tod_event_versions', 'dial_split_df', 'program_week_df', 'program_day_df', 'controller_id'}
methods: {'get_controller_spat_period', 'add_node', 'get_tod_event_time_splits', 'get_tod_event_starts'}
Get corridor SPaT information for a time period over a range of days#
[8]:
# specify date range
date_range = ["2022-03-07", "2022-03-08", "2022-03-09", "2022-03-10", "2022-03-11"]
# corridor name
name = 'adams_rd'
# specify nodes in the corridor
node_ls = ["62183487", "62300620", "61950324", "62344947", "62344935", "62355467", "62036287"]
# get TOD splits based on all the TOD boundaries from the nodes in this node list
tod_splits = region_spat.get_tod_event_time_splits(date_range, node_ls)
print(tod_splits)
[[0.0, 6.0], [6.0, 7.0], [7.0, 10.0], [10.0, 15.0], [15.0, 15.25], [15.25, 15.667], [15.667, 19.0], [19.0, 22.0], [22.0, 24]]
[9]:
# examine the first tod split
tod_split = tod_splits[0]
# get the corridor spat
corridor_period_spat = region_spat.get_corridor_spat_period(name, node_ls, date_range, tod_split)
WARNING: nodes have different cycle lengths for date range ['2022-03-07', '2022-03-08', '2022-03-09', '2022-03-10', '2022-03-11'] over time interval [0.0, 6.0]
{
"90": [
"62183487",
"62344947",
"62344935",
"62355467"
],
"0": [
"62300620",
"61950324",
"62036287"
]
}
A warning is raised because not all of the nodes have the same cycle length. Some of the nodes in this example have a flashing operation during the night hours, so the cycle is 0. This is also stored in the object.
[10]:
print(corridor_period_spat.same_cycle_lengths)
False
Now let’s get a corridor period when all the nodes have the same cycle length:
[11]:
# examine the third tod split
tod_split = tod_splits[2]
# get the corridor spat
corridor_period_spat = region_spat.get_corridor_spat_period(name, node_ls, date_range, tod_split)
# check that cycle length is the same
print(corridor_period_spat.same_cycle_lengths)
True
[12]:
show_attributes_and_methods(corridor_period_spat)
attributes: {'controller_spat_period_dict', 'timezone', 'date_ls', 'controller_cycle_lengths', 'corridor_id', 'node_id_ls', 'cycle_length', 'node_controller_dict', 'node_offsets', 'time_range', 'same_cycle_lengths'}
methods: {'get_movement_spat_period', 'get_controller_spat_period'}
[13]:
corridor_period_spat.node_offsets
[13]:
{'62183487': 37,
'62300620': 40,
'61950324': 89,
'62344947': 35,
'62344935': 35,
'62355467': 78,
'62036287': 69}
Get the period SPaT information for a node in the corridor#
[14]:
intersection_id = corridor_period_spat.node_id_ls[0]
controller_period_spat = corridor_period_spat.get_controller_spat_period(intersection_id)
show_attributes_and_methods(controller_period_spat)
attributes: {'next_phases', 'clearance_intervals', 'yellow_intervals', 'shift', 'controller_id', 'version_history', 'phase_mappings', 'ring_structure_df', 'movement_spat_period_dict', 'fixed_ends', 'date_ls', 'tod_events', 'phase_modes', 'node_id_ls', 'program_weeks', 'ring_phases', 'ped_channels', 'dial_splits', 'cycle', 'timezone', 'movement_mappings', 'fixed_starts', 'concurrent_phases', 'phase_starts', 'offset', 'program_days', 'channels', 'phase_splits', 'min_splits', 'time_range'}
methods: {'get_movement_spat_period'}
Get the period SPaT information for a movement in the corridor#
[16]:
# examine the SB movement
movement_index = 8
You can either retrieve the movement information from either the corridor or controller level
[17]:
movement_period_spat_from_corridor = corridor_period_spat.get_movement_spat_period(node_id, movement_index)
movement_period_spat_from_controller = controller_period_spat.get_movement_spat_period(node_id, movement_index)
movement_period_spat_from_corridor == movement_period_spat_from_controller
WARNING: node id 61950324 not in controller 279
[17]:
False
[19]:
movement_period_spat = movement_period_spat_from_corridor
movement_period_spat.__dict__
[19]:
{'movement_id': None,
'movement_index': 8,
'phase_ls': [2],
'fx_end': True,
'fx_start': True,
'green_start': 89,
'green_duration': 60,
'yellow': 3.5,
'clearance': 2.5}
The movement id needs to be matched from the map data
Get intersection SPaT information for a time period over a range of days#
You can also get an intersection’s SPaT information for a time period directly from the controller history
[20]:
tod_splits = controller_spat.get_tod_event_time_splits(date_range)
print(tod_splits)
[[0.0, 7.0], [7.0, 10.0], [10.0, 15.0], [15.0, 15.25], [15.25, 15.667], [15.667, 19.0], [19.0, 22.0], [22.0, 24]]
It is recommended that you call the controller SPaT period within these tod splits. Otherwise an error might be raised because of incompatible parameters.
[21]:
# try a different tod split
tod_split = [6, 8]
controller_period_spat = controller_spat.get_controller_spat_period(date_range, tod_split)
WARNING: time interval includes more than one programmed SPaT TOD on 2022-03-07 for controller 348 (nodes: ['61950324'])
[
[
0.0,
7.0
],
[
7.0,
10.0
]
]
WARNING: time interval includes more than one programmed SPaT TOD on 2022-03-08 for controller 348 (nodes: ['61950324'])
[
[
0.0,
7.0
],
[
7.0,
10.0
]
]
WARNING: time interval includes more than one programmed SPaT TOD on 2022-03-09 for controller 348 (nodes: ['61950324'])
[
[
0.0,
7.0
],
[
7.0,
10.0
]
]
WARNING: time interval includes more than one programmed SPaT TOD on 2022-03-10 for controller 348 (nodes: ['61950324'])
[
[
0.0,
7.0
],
[
7.0,
10.0
]
]
WARNING: time interval includes more than one programmed SPaT TOD on 2022-03-11 for controller 348 (nodes: ['61950324'])
[
[
0.0,
7.0
],
[
7.0,
10.0
]
]
WARNING: date range includesF more than one dial splits for controller 348 (nodes: ['61950324'])
{
"5:5": [
"2022-03-07_0.0-7.0",
"2022-03-08_0.0-7.0",
"2022-03-09_0.0-7.0",
"2022-03-10_0.0-7.0",
"2022-03-11_0.0-7.0"
],
"2:1": [
"2022-03-07_7.0-10.0",
"2022-03-08_7.0-10.0",
"2022-03-09_7.0-10.0",
"2022-03-10_7.0-10.0",
"2022-03-11_7.0-10.0"
]
}
---------------------------------------------------------------------------
Exception Traceback (most recent call last)
Cell In[21], line 4
1 # try a different tod split
2 tod_split = [6, 8]
----> 4 controller_period_spat = controller_spat.get_controller_spat_period(date_range, tod_split)
File d:\mtldp_project\mtldp-meta-utils\mtldp\meta\SPaT\HistoryControllerSPaT.py:125, in HistoryControllerSPaT.get_controller_spat_period(self, date_ls, time_range)
122 program_days = self._get_program_days(program_weeks)
123 dial_splits, tod_events = self._get_dial_splits_and_tod_events(program_days, time_range,
124 version_history['tod_events'])
--> 125 parameter_set = self._get_timing_parameters(dial_splits, version_history['dial_splits'], time_range)
126 shift = self._get_shift(time_range, parameter_set)
127 return PeriodControllerSPaT(self.controller_id, self.node_id_ls, time_range, ring_structure_df,
128 self.phase_mappings, self.movement_mappings, date_ls, version_history,
129 program_weeks, program_days, dial_splits, tod_events, parameter_set, shift)
File d:\mtldp_project\mtldp-meta-utils\mtldp\meta\SPaT\HistoryControllerSPaT.py:317, in HistoryControllerSPaT._get_timing_parameters(self, dial_splits, dial_split_versions, time_range)
313 parameter_set[dial_split] = {"phase_splits": dial_split_info.phase_splits,
314 "cycle": dial_split_info.cycle, "offset": dial_split_info.offset,
315 "phase_modes": dial_split_info.phase_modes}
316 final_parameter_set = parameter_set[dial_split]
--> 317 self._uniform_info(parameter_set,
318 note=f'timing parameter set (phase splits, cycle, offset, phase modes) for time interval '
319 f'{time_range}', fatal=True)
320 return final_parameter_set
File d:\mtldp_project\mtldp-meta-utils\mtldp\meta\SPaT\HistoryControllerSPaT.py:342, in HistoryControllerSPaT._uniform_info(self, a_dict, note, fatal)
340 if len(group_dict) > 1:
341 if fatal:
--> 342 raise Exception(f"FATAL: date range includes more than one {note} for controller {self.controller_id} "
343 f"(nodes: {self.node_id_ls}) \n"
344 f"{json.dumps(group_dict, indent=1)}")
345 else:
346 print(f"WARNING: date range includes more than one {note} for controller {self.controller_id} "
347 f"(nodes: {self.node_id_ls}) \n"
348 f"{json.dumps(group_dict, indent=1)}")
Exception: FATAL: date range includes more than one timing parameter set (phase splits, cycle, offset, phase modes) for time interval [6, 8] for controller 348 (nodes: ['61950324'])
{
"{'phase_splits': {1: -1, 2: 0, 3: -1, 4: 0, 5: -1, 6: -1, 7: -1, 8: -1}, 'cycle': 0, 'offset': 0, 'phase_modes': {1: -1, 2: 8, 3: -1, 4: 9, 5: -1, 6: -1, 7: -1, 8: -1}}": [
"5:5"
],
"{'phase_splits': {1: -1, 2: 60, 3: -1, 4: 30, 5: -1, 6: -1, 7: -1, 8: -1}, 'cycle': 90, 'offset': 89, 'phase_modes': {1: -1, 2: 1, 3: -1, 4: 7, 5: -1, 6: -1, 7: -1, 8: -1}}": [
"2:1"
]
}
In this case, there are different cycle lengths within this time range, so 6:00 AM to 8:00 AM is not a viable TOD split for this date range.
[22]:
# try a TOD split within the ones generated before
tod_split = [8.5, 9.5]
controller_period_spat = controller_spat.get_controller_spat_period(date_range, tod_split)
[23]:
show_attributes_and_methods(controller_period_spat)
attributes: {'next_phases', 'clearance_intervals', 'yellow_intervals', 'shift', 'controller_id', 'version_history', 'phase_mappings', 'ring_structure_df', 'movement_spat_period_dict', 'fixed_ends', 'date_ls', 'tod_events', 'phase_modes', 'node_id_ls', 'program_weeks', 'ring_phases', 'ped_channels', 'dial_splits', 'cycle', 'timezone', 'movement_mappings', 'fixed_starts', 'concurrent_phases', 'phase_starts', 'offset', 'program_days', 'channels', 'phase_splits', 'min_splits', 'time_range'}
methods: {'get_movement_spat_period'}
[24]:
controller_period_spat.phase_starts
[24]:
{2: 0, 4: 60}