-
Notifications
You must be signed in to change notification settings - Fork 3
/
geo_clustering.py
246 lines (216 loc) · 11.8 KB
/
geo_clustering.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
import collections
import itertools
import math
from abc import abstractmethod, ABC
from typing import List, Tuple, Iterator, Optional
import numpy as np
import sklearn.cluster
from .geo_coords import GeoCoord
from .local_coords import LocalCoordinateSystem
from ..clustering import GreedyAgglomerativeClustering
class GeoCoordClusterer(ABC):
@abstractmethod
def fit_geo_coords(self, geo_coords: List[GeoCoord]):
"""
:param geo_coords: the coordinates to be clustered
"""
pass
@abstractmethod
def clusters_indices(self) -> Tuple[List[List[int]], List[int]]:
"""
:return: a tuple (clusters, outliers), where clusters is a list of point indices, one list for each
cluster containing the indices of points within the cluster, and outliers is the list of indices of points not within
clusters
"""
pass
class GreedyAgglomerativeGeoCoordClusterer(GeoCoordClusterer):
def __init__(self,
max_min_distance_for_merge_m: float,
max_distance_m: float,
min_cluster_size: int,
lcs: LocalCoordinateSystem = None):
"""
:param max_min_distance_for_merge_m: the maximum distance, in metres, for the minimum distance between two existing clusters for a merge
to be admissible
:param max_distance_m: the maximum distance, in metres, between any two points for the points to be allowed to be in the same cluster
:param min_cluster_size: the minimum number of points any valid cluster must ultimately contain; the points in any smaller clusters
shall be considered as outliers
:param lcs: the local coordinate system to use for clustering; if None, compute based on mean coordinates passed when fitting
"""
self.lcs = lcs
self.min_cluster_size = min_cluster_size
self.max_min_distance_for_merge = max_min_distance_for_merge_m
self.max_distance_m = max_distance_m
self.squared_max_min_distance_for_merge = max_min_distance_for_merge_m * max_min_distance_for_merge_m
self.squared_max_distance = max_distance_m * max_distance_m
self.local_points = None
self.m_min_distance: Optional["GreedyAgglomerativeGeoCoordClusterer.Matrix"] = None
self.m_max_squared_distance: Optional["GreedyAgglomerativeGeoCoordClusterer.Matrix"] = None
self.clusters = None
class Matrix:
UNSET_VALUE = np.inf
def __init__(self, dim: int):
self.m = np.empty((dim, dim))
self.m.fill(np.inf)
def set(self, c1: int, c2: int, value: float):
self.m[c1][c2] = value
self.m[c2][c1] = value
def get(self, c1: int, c2: int) -> float:
return self.m[c1][c2]
class LocalPoint:
def __init__(self, xy: np.ndarray, idx: int):
self.idx = idx
self.xy = xy
class Cluster(GreedyAgglomerativeClustering.Cluster):
def __init__(self, point: "GreedyAgglomerativeGeoCoordClusterer.LocalPoint", idx: int,
clusterer: 'GreedyAgglomerativeGeoCoordClusterer'):
self.idx = idx
self.clusterer = clusterer
self.points = [point]
def merge_cost(self, other: "GreedyAgglomerativeGeoCoordClusterer.Cluster"):
cartesian_product = itertools.product(self.points, other.points)
min_squared_distance = math.inf
max_squared_distance = 0
for p1, p2 in cartesian_product:
diff = p1.xy - p2.xy
squared_distance = np.dot(diff, diff)
if squared_distance > self.clusterer.squared_max_distance:
max_squared_distance = math.inf
break
else:
min_squared_distance = min(squared_distance, min_squared_distance)
# fill cache: the max value takes precedence; if it is inf (no merge admissible), then the min value is also set to inf;
# the min value valid only if the max value is finite
self.clusterer.m_max_squared_distance.set(self.idx, other.idx, max_squared_distance)
if np.isinf(max_squared_distance):
self.clusterer.m_min_distance.set(self.idx, other.idx, np.inf)
else:
self.clusterer.m_min_distance.set(self.idx, other.idx, min_squared_distance)
if np.isinf(max_squared_distance):
return math.inf
if min_squared_distance <= self.clusterer.squared_max_min_distance_for_merge:
return min_squared_distance
return math.inf
def merge(self, other):
self.points += other.points
def fit_geo_coords(self, geo_coords: List[GeoCoord]) -> None:
self.m_min_distance = self.Matrix(len(geo_coords))
self.m_max_squared_distance = self.Matrix(len(geo_coords))
if self.lcs is None:
mean_coord = GeoCoord.mean_coord(geo_coords)
self.lcs = LocalCoordinateSystem(mean_coord.lat, mean_coord.lon)
self.local_points = [self.LocalPoint(np.array(self.lcs.get_local_coords(p.lat, p.lon)), idx) for idx, p in enumerate(geo_coords)]
clusters = [self.Cluster(lp, i, self) for i, lp in enumerate(self.local_points)]
gac = GreedyAgglomerativeClustering(clusters,
merge_candidate_determination_strategy=self.MergeCandidateDeterminationStrategy(self.max_distance_m, self))
clusters = gac.apply_clustering()
self.clusters = clusters
def clusters_indices(self) -> Tuple[List[List[int]], List[int]]:
outliers = []
clusters = []
for c in self.clusters:
indices = [p.idx for p in c.points]
if len(c.points) < self.min_cluster_size:
outliers.extend(indices)
else:
clusters.append(indices)
return clusters, outliers
class MergeCandidateDeterminationStrategy(GreedyAgglomerativeClustering.MergeCandidateDeterminationStrategy):
def __init__(self, search_radius_m: float, parent: "GreedyAgglomerativeGeoCoordClusterer"):
super().__init__()
self.parent = parent
self.searchRadiusM = search_radius_m
def set_clusterer(self, clusterer: GreedyAgglomerativeClustering):
super().set_clusterer(clusterer)
points = []
for wc in self.clusterer.wrapped_clusters:
c: GreedyAgglomerativeGeoCoordClusterer.Cluster = wc.cluster
for p in c.points:
points.append(p.xy)
assert len(points) == len(self.clusterer.wrapped_clusters)
points = np.stack(points)
self.kdtree = sklearn.neighbors.KDTree(points)
def iter_candidate_indices(self, wc: "GreedyAgglomerativeClustering.WrappedCluster", initial: bool,
merged_cluster_indices: Tuple[int, int] = None) -> Iterator[int]:
c: GreedyAgglomerativeGeoCoordClusterer.Cluster = wc.cluster
if initial:
local_point = c.points[0] # pick any point from wc, since we use maximum cluster extension as search radius
indices = self.kdtree.query_radius(np.reshape(local_point.xy, (1, 2)), self.searchRadiusM)[0]
candidate_set = set()
for idx in indices:
wc = self.clusterer.wrapped_clusters[idx]
candidate_set.add(wc.get_cluster_association().idx)
yield from sorted(candidate_set)
else:
# The new distance values (max/min) between wc and any cluster index otherIdx can be computed from the cached distance
# values of the two clusters from which wc was created through a merge:
# The max distance is the maximum of the squared distances of the original clusters (and if either is inf, then
# a merge is definitely inadmissible, because one of the original clusters was already too far away).
# The min distance is the minimum of the squred distances of the original clusters.
c1, c2 = merged_cluster_indices
max1 = self.parent.m_max_squared_distance.m[c1]
max2 = self.parent.m_max_squared_distance.m[c2]
max_combined = np.maximum(max1, max2)
for otherIdx, maxSqDistance in enumerate(max_combined):
min_sq_distance = np.inf
if maxSqDistance <= self.parent.squared_max_distance:
wc_other = self.clusterer.wrapped_clusters[otherIdx]
if wc_other.is_merged():
continue
min1 = self.parent.m_min_distance.get(c1, otherIdx)
min2 = self.parent.m_min_distance.get(c2, otherIdx)
min_sq_distance = min(min1, min2)
if min_sq_distance <= self.parent.squared_max_min_distance_for_merge:
yield GreedyAgglomerativeClustering.ClusterMerge(wc, wc_other, min_sq_distance)
# update cache
self.parent.m_max_squared_distance.set(wc.idx, otherIdx, maxSqDistance)
self.parent.m_min_distance.set(wc.idx, otherIdx, min_sq_distance)
class SkLearnGeoCoordClusterer(GeoCoordClusterer):
def __init__(self, clusterer, lcs: LocalCoordinateSystem = None):
"""
:param clusterer: a clusterer from sklearn.cluster
:param lcs: the local coordinate system to use for Euclidian conversion; if None, determine from data (using mean coordinate as
centre)
"""
self.lcs = lcs
self.clusterer = clusterer
self.local_points = None
def fit_geo_coords(self, geo_coords: List[GeoCoord]):
if self.lcs is None:
mean_coord = GeoCoord.mean_coord(geo_coords)
self.lcs = LocalCoordinateSystem(mean_coord.lat, mean_coord.lon)
self.local_points = [self.lcs.get_local_coords(p.lat, p.lon) for p in geo_coords]
self.clusterer.fit(self.local_points)
def _clusters(self, mode):
clusters = collections.defaultdict(list)
outliers = []
for idxPoint, idxCluster in enumerate(self.clusterer.labels_):
if mode == "localPoints":
item = self.local_points[idxPoint]
elif mode == "indices":
item = idxPoint
else:
raise ValueError()
if idxCluster >= 0:
clusters[idxCluster].append(item)
else:
outliers.append(item)
return list(clusters.values()), outliers
def clusters_local_points(self) -> Tuple[List[List[Tuple[float, float]]], List[Tuple[float, float]]]:
"""
:return: a tuple (clusters, outliers), where clusters is a dictionary mapping from cluster index to
the list of local points within the cluster and outliers is a list of local points not within
clusters
"""
return self._clusters("localPoints")
def clusters_indices(self) -> Tuple[List[List[int]], List[int]]:
return self._clusters("indices")
class DBSCANGeoCoordClusterer(SkLearnGeoCoordClusterer):
def __init__(self, eps, min_samples, lcs: LocalCoordinateSystem = None, **kwargs):
"""
:param eps: the maximum distance between two samples for one to be considered as in the neighbourhood of the other
:param min_samples: the minimum number of samples that must be within a neighbourhood for a cluster to be formed
:param lcs: the local coordinate system for conversion to a Euclidian space
:param kwargs: additional arguments to pass to DBSCAN (see https://scikit-learn.org/stable/modules/generated/sklearn.cluster.DBSCAN.html)
"""
super().__init__(sklearn.cluster.DBSCAN(eps=eps, min_samples=min_samples, **kwargs), lcs)