-
Notifications
You must be signed in to change notification settings - Fork 0
/
4_spatial_tensors.py
318 lines (252 loc) · 11.4 KB
/
4_spatial_tensors.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
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
from STM import SpeedTransitionMatrix
from misc import database, config
from misc.misc import plot_heatmap, save_pickle_data, get_time
import numpy as np
from scipy.spatial import distance
import math
import pandas as pd
import tensorly as ty
from tensorly.decomposition import non_negative_parafac
def create_coordinate_matrix(sp, xn, yn, lons, lats):
"""
Creates xn times yn matrix of GNSS points.
:param sp: Starting GNSS point.
:param xn: Number of rectangles (columns).
:param yn: Number of rectangles (rows).
:param lons: Longitude step.
:param lats: Latitude step.
:return: Matrix of GNSS points for rectangle drawing. Every cell consists of a tuple with four points (lon1, lat1, lon2, lat2, lon3, lat3, lon4, lat4)
"""
coordinate_matrix = []
column_values = []
for ii in range(1, yn + 1):
for jj in range(1, xn + 1):
lon1 = sp[0] + ((jj - 1) * lons)
lat1 = sp[1] - ((ii - 1) * lats)
lon2 = sp[0] + (jj * lons)
lat2 = sp[1] - (ii * lats)
lon3 = lon1 + lons
lat3 = lat1
lon4 = lon2 - lons
lat4 = lat2
column_values.append((lon1, lat1, lon2, lat2, lon3, lat3, lon4, lat4))
coordinate_matrix.append(column_values)
column_values = []
return coordinate_matrix
def get_mass_center(m):
max_val = 0.2 * np.max(m) # Filter: remove 20% of maximal value.
m = np.where(m < max_val, 0, m)
m = m / np.sum(m)
# marginal distributions
dx = np.sum(m, 1)
dy = np.sum(m, 0)
# expected values
X, Y = m.shape
cx = np.sum(dx * np.arange(X))
cy = np.sum(dy * np.arange(Y))
return int(cx), int(cy)
def diag_dist(point):
# Max distance to the diagonal (square matrix m x m) is: diagonal_length / 2.
max_d = (config.MAX_INDEX * math.sqrt(2)) / 2
distan = []
for d in config.DIAG_LOCS:
distan.append(distance.euclidean(d, point))
return round(min(distan) / max_d * 100, 2) # Relative distance.
# def from_max_distance(point):
# max_point = (config.MAX_INDEX, config.MAX_INDEX)
# origin = (0, 0)
# max_d = distance.euclidean(origin, max_point)
# d = round(distance.euclidean(max_point, point) / max_d * 100, 2)
# return d
def from_origin_distance(point):
max_point = (config.MAX_INDEX, config.MAX_INDEX)
origin = (0, 0)
max_d = distance.euclidean(origin, max_point)
d = round(distance.euclidean(origin, point) / max_d * 100, 2)
return d
def get_link_ids_square(points, link_info):
try:
link_ids = link_info[(link_info.x_b > points[0][0])
& (link_info.y_b < points[0][1])
& (link_info.x_b < points[1][0])
& (link_info.y_b > points[1][1])]
if len(link_ids.link_id.values) == 0:
return None
return link_ids.link_id.values
except:
return None
print('Script {0} started ... '.format(__file__))
t_start = get_time()
config.initialize_paths()
config.initialize_db_setup()
config.initialize_stm_setup()
db, client = database.init('SpeedTransitionDB')
#col_name = "spatialMatrixRWLNEWrel"
col_name = config.SM_COLLECTION+'rel'
tensor_col_name = config.TENSOR_COLLECTION
tensor_rank = 10
spatial_square = dict({})
total_data = list([])
lon_step = 0.006545 # ~500[m]
lat_step = 0.004579 # ~500[m]
x_num = 50 # Number of rectangles (columns).
y_num = 20 # Number of rectangles (rows).
lon_start = 15.830326
lat_start = 45.827299
start_point = (lon_start, lat_start)
coordinate_matrix = create_coordinate_matrix(sp=start_point,
xn=x_num,
yn=y_num,
lons=lon_step,
lats=lat_step)
info = pd.read_csv(r'links_info.csv', sep=';')
###############################################
all_stms = []
###############################################
none_counter = 0
total_counter = 0
for i in range(0, len(coordinate_matrix)):
for j in range(0, len(coordinate_matrix[0])):
print("i=%d\t\tj=%d" % (i, j))
total_counter += 1
p1 = (coordinate_matrix[i][j][0], coordinate_matrix[i][j][1])
p2 = (coordinate_matrix[i][j][2], coordinate_matrix[i][j][3])
links_inside = get_link_ids_square(points=(p1, p2), link_info=info)
if links_inside is not None:
c = 0
frontal_slices = []
valid_transitions = []
temp = []
temp_tran = []
try:
n_intervals = 8
for interval in range(0, n_intervals):
for link in links_inside:
# transitions = database.selectSome(db, col_name, {'$or': [{'origin_id': int(link)}, {'destination_id': int(link)}]})
transitions = database.selectSome(db, col_name, {'origin_id': int(link)})
for tran in transitions:
matrix = np.array(tran['intervals'][interval]['winter']['working'])
if int(np.sum(matrix)) > 20:
######################################################
cx, cy = get_mass_center(matrix)
dist_diagonal = diag_dist(point=(cx, cy))
dist_from_origin = from_origin_distance(point=(cx, cy))
anomaly = False
if dist_diagonal >= 46:
anomaly = True
traff_state = 0
if dist_from_origin > 67:
traff_state = 0
elif 40 < dist_from_origin < 67:
traff_state = 1
else:
traff_state = 2
all_stms.append({'stm': matrix,
'interval': interval,
'season': 'winter',
'day': 'working',
'com_position': [cx, cy],
'com_diag_dist': dist_diagonal,
'dist_from_origin': dist_from_origin,
'traff_state': traff_state,
'anomaly': anomaly
})
######################################################
temp.append(list(matrix.flatten()))
c += 1
# temp_tran.append((tran['origin_id'], tran['destination_id']))
valid_transitions.append((tran['origin_id'], tran['destination_id']))
# temp = np.array(temp).reshape((400, len(temp)))
frontal_slices.append(temp)
# valid_transitions.append(temp_tran)
temp = []
temp_tran = []
except:
print('Warning: There are no transitions with oringin_id: %s' % link)
slices_length = [len(slice) for slice in frontal_slices]
# print(slices_length)
n_trans = min(slices_length)
if n_trans == 0:
continue
# print()
# valid_transitions = [x[0:n_trans] for x in valid_transitions]
valid_transitions = valid_transitions[0:n_trans]
tensor = np.zeros((400, n_trans, 8))
for f_slice_id in range(0, len(frontal_slices)):
for matrix_id in range(0, len(frontal_slices[f_slice_id])):
if matrix_id >= n_trans:
continue
tensor[:, matrix_id, f_slice_id] = frontal_slices[f_slice_id][matrix_id]
factors = non_negative_parafac(tensor=ty.tensor(tensor), rank=tensor_rank, verbose=0)
spatial_square = dict({})
spatial_square['p1'] = p1
spatial_square['p2'] = p2
spatial_square['tensor'] = tensor
spatial_square['links_inside'] = links_inside
spatial_square['valid_transitions'] = valid_transitions
spatial_square['xy_position'] = [i, j]
spatial_square['char_matrices'] = list([])
spatial_square['spatial_matrix'] = factors.factors[1].tolist()
spatial_square['temporal_matrix'] = factors.factors[2].tolist()
factor_index = 0
for column in range(0, factors.factors[0].shape[1]):
orig = factors.factors[0][:, column].reshape(20, 20)
rounded = orig / np.sum(orig)
rounded = np.round(rounded, decimals=2)
cx, cy = get_mass_center(orig)
dist_diagonal = diag_dist(point=(cx, cy))
dist_from_origin = from_origin_distance(point=(cx, cy))
anomaly = False
if dist_diagonal >= 46:
anomaly = True
traff_state = 0
if dist_from_origin > 67:
traff_state = 0
elif 40 < dist_from_origin < 67:
traff_state = 1
else:
traff_state = 2
chm = {'orig': orig.tolist(),
'rounded': rounded.tolist(),
'com_position': [cx, cy],
'com_diag_dist': dist_diagonal,
'dist_from_origin': dist_from_origin,
'traff_state': traff_state,
'factor_id': factor_index,
'anomaly': anomaly,
'class': 0
}
if anomaly:
sm = factors.factors[1][:, factor_index].tolist()
spatial_max_id = sm.index(max(sm))
tm = factors.factors[2][:, factor_index].tolist()
temporal_max_id = tm.index(max(tm))
chm['max_spatial_id'] = spatial_max_id
chm['spatial_anomaly_char'] = sm
chm['anomalous_trans'] = valid_transitions[spatial_max_id]
chm['max_temporal_id'] = temporal_max_id
chm['temporal_anomaly_char'] = tm
spatial_square['char_matrices'].append(chm)
factor_index += 1
total_data.append(spatial_square)
else:
none_counter += 1
spatial_square = dict({})
spatial_square['p1'] = p1
spatial_square['p2'] = p2
spatial_square['tensor'] = None
spatial_square['links_inside'] = None
spatial_square['xy_position'] = [i, j]
spatial_square['char_matrices'] = None
total_data.append(spatial_square)
# TODO: spatial_square insert into database
########################################################
save_pickle_data('all_matrices.pkl', all_stms)
########################################################
# t1 = get_time()
# save_pickle_data('spatialTensors5.pkl', total_data)
# t2 = get_time()
# print('Pickle save time: {0}'.format(t2 - t1))
#
# t_end = get_time()
# print('Exe time: {0}'.format(t_end - t_start))