-
Notifications
You must be signed in to change notification settings - Fork 4
/
trajectory.py
276 lines (217 loc) · 8.44 KB
/
trajectory.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
import math
import numpy as np
import pandas as pd
import osmnx as ox
import geopandas as gpd
import networkx as nx
from numba import jit
from pyquadkey2 import quadkey
from geo.qk import tile_to_str
from raster.drawing import smooth_line
from itertools import pairwise
from db.api import EVedDb
def geocode_address(address, crs=4326):
geocode = gpd.tools.geocode(address,
provider='nominatim',
user_agent="QuadKey trajectory query").to_crs(crs)
return geocode.iloc[0].geometry.y, geocode.iloc[0].geometry.x
def get_qk_line(loc0, loc1, level):
qk0 = quadkey.from_geo((loc0['y'], loc0['x']), level)
qk1 = quadkey.from_geo((loc1['y'], loc1['x']), level)
((tx0, ty0), _) = qk0.to_tile()
((tx1, ty1), _) = qk1.to_tile()
line = smooth_line(tx0, ty0, tx1, ty1)
return [(quadkey.from_str(tile_to_str(int(p[0]), int(p[1]), level)), p[2]) for p in line if p[2] > 0.0]
def load_signal_range(r):
db = EVedDb()
sql = """
select match_latitude
, match_longitude
from signal
where signal_id >= ? and signal_id <= ?
"""
points = []
all_points = db.query(sql, [int(r[0]), int(r[1])])
for p0, p1 in pairwise(all_points):
if len(points) == 0:
points.append(p0)
elif p0 != p1:
points.append(p1)
return points
@jit(nopython=True)
def get_contiguous_ranges(signal_ini, signal_end):
ranges = np.zeros((signal_ini.shape[0], 2))
ini = signal_ini[0]
end = signal_end[0]
j = 0
for i in range(1, signal_ini.shape[0]):
end = signal_end[i - 1]
if signal_ini[i] != end:
ranges[j, 0] = ini
ranges[j, 1] = end
j += 1
ini = signal_ini[i]
if j == 0:
ranges[j, 0] = ini
ranges[j, 1] = end
j += 1
return ranges[:j, :]
def load_trajectory_quadkeys(traj_id):
db = EVedDb()
sql = """
select s.quadkey
from signal s
inner join trajectory t on s.vehicle_id = t.vehicle_id and s.trip_id = t.trip_id
where t.traj_id = ?;
"""
qks = {qk[0] for qk in db.query(sql, [traj_id])}
return qks
def load_trajectory_points(traj_id, unique=False):
db = EVedDb()
distinct = "distinct" if unique else ""
sql = f"""
select {distinct}
s.match_latitude
, s.match_longitude
, s.bearing
from signal s
inner join trajectory t on s.vehicle_id = t.vehicle_id and s.trip_id = t.trip_id
where t.traj_id = ?
order by s.time_stamp;
"""
return db.query(sql, [traj_id])
def load_link_points(link_id):
db = EVedDb()
get_range_sql = "select signal_ini, signal_end from link where link_id=?"
ranges = db.query(get_range_sql, [link_id])
if len(ranges):
get_points_sql = """
select distinct match_latitude
, match_longitude
from signal
where signal_id >= ? and signal_id <= ?
order by signal_id
"""
return db.query(get_points_sql, [ranges[0][0], ranges[0][1]])
else:
return []
def jaccard_similarity(set0, set1):
return len(set0 & set1) / len(set0 | set1)
class GraphRoute(object):
def __init__(self, graph):
self.graph = graph
self.graph = ox.add_edge_speeds(self.graph)
self.graph = ox.add_edge_travel_times(self.graph)
self.graph = ox.bearing.add_edge_bearings(self.graph)
self.route = None
def generate_route(self, addr_ini, addr_end, weight='travel_time'):
g = self.graph
loc_ini = geocode_address(addr_ini)
loc_end = geocode_address(addr_end)
node_ini = ox.distance.nearest_nodes(g, loc_ini[1], loc_ini[0])
node_end = ox.distance.nearest_nodes(g, loc_end[1], loc_end[0])
self.route = nx.shortest_path(g, node_ini, node_end, weight=weight)
return self.route
def has_route(self):
return self.route is not None
def get_route_nodes(self):
return [self.graph.nodes[n] for n in self.route]
def get_route_quadkeys(self, level=20):
qks = set()
g = self.graph
shift = 64 - 2 * level
for n0, n1 in pairwise(self.route):
edge = g[n0][n1]
l0 = g.nodes[n0]
l1 = g.nodes[n1]
qks.update([(qk.to_quadint() >> shift, edge[0]['bearing'])
for qk, _ in get_qk_line(l0, l1, level)])
return list(qks)
def get_overlapping_links(self, level=20, angle_delta=2.5):
qks = self.get_route_quadkeys(level)
cos_angle_delta = math.cos(math.radians(angle_delta))
sql = """
select q.link_id
, l.traj_id
, l.signal_ini
, l.signal_end
from link_qk q
inner join link l on l.link_id = q.link_id
where q.quadkey = ? and l.bearing > 0 and cos(radians(l.bearing - ?)) >= ?;
"""
db = EVedDb()
links = set()
for qk, bearing in qks:
links.update(db.query(sql, [qk, bearing, cos_angle_delta]))
return np.array(list(links))
def get_matching_trajectories(self, level=20, angle_delta=2.5):
links = self.get_overlapping_links(level, angle_delta)
trajectories = np.unique(links[:, 1])
return trajectories, links
def get_overlapping_signal_ranges(self, level=20, angle_delta=2.5):
trajectories, links = self.get_matching_trajectories(level, angle_delta)
ranges = []
for t in trajectories:
index = links[:, 1] == t
signal_ini = links[index, 2]
signal_end = links[index, 3]
ranges.extend(get_contiguous_ranges(signal_ini, signal_end).tolist())
return ranges
def calculate_trajectory_matches(self, level=20):
trajectories, links = self.get_matching_trajectories(level)
route_qks = {qk[0] for qk in self.get_route_quadkeys(level)}
data = []
for trajectory in trajectories:
traj_qks = load_trajectory_quadkeys(int(trajectory))
similarity = jaccard_similarity(traj_qks, route_qks)
data.append((trajectory, similarity))
return data
def get_top_match_trajectories(self, level=20, top=0.05):
match_df = pd.DataFrame(data=self.calculate_trajectory_matches(level), columns=['traj_id', 'similarity'])
match_df["percent_rank"] = match_df["similarity"].rank(pct=True)
filtered_df = match_df[match_df["percent_rank"] > (1.0 - top)]
trajectories = filtered_df["traj_id"].values
return trajectories
def load_matching_links(traj_id, angle_delta=2.5):
db = EVedDb()
sql = """
select q.link_id
, q.quadkey
, l.traj_id
from link_qk q
inner join link l on l.link_id = q.link_id
inner join (
select qk.quadkey
, lk.bearing
from link_qk qk
inner join link lk on lk.link_id = qk.link_id
where lk.traj_id = ?
) x on x.quadkey = q.quadkey
where l.bearing > 0 and x.bearing > 0 and cos(radians(x.bearing - l.bearing)) >= cos(radians(?));
"""
traj_df = db.query_df(sql, [traj_id, angle_delta])
return traj_df
class GraphTrajectory(object):
def __init__(self, traj_id):
assert isinstance(traj_id, int)
self.traj_id = int(traj_id)
def get_top_matching_trajectories(self, top=0.05, exclude_self=True):
df = load_matching_links(self.traj_id)
query_set = set(df[df["traj_id"] == self.traj_id]["quadkey"].values)
if exclude_self:
df = df[df["traj_id"] != self.traj_id]
trajectories = np.unique(df["traj_id"].values)
traj_df = pd.DataFrame(data=trajectories, columns=["traj_id"])
traj_df["similarity"] = [jaccard_similarity(query_set,
set(df[df["traj_id"] == t]["quadkey"].values))
for t in trajectories]
traj_df["percent_rank"] = traj_df["similarity"].rank(pct=True)
filtered_df = traj_df[traj_df["percent_rank"] > (1.0 - top)]
trajectories = filtered_df["traj_id"].values
return trajectories
def get_matching_links(self, exclude_self=True):
df = load_matching_links(self.traj_id)
if exclude_self:
df = df[df["traj_id"] != self.traj_id]
links = np.unique(df["link_id"].values)
return links