trajectoryClustering/cluster.py at main · fillif3/trajectoryClustering · GitHub

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
import math
import numpy as np
import copy

class cluster:

    def __init__(self, k, beta):
        self.trajectories = []
        self.trajectoriesProb = []
        self.trajectoriesIndex = []
        self.variance = [''] * k
        self.mean = [''] * k
        self.k = k
        self.beta = beta

    def probabilityOfTrajInCluster(self, traj):
        LogProb = 0
        # beta =len(traj)/self.k
        for i in range(len(traj)):

            index = math.floor(i / self.beta)
            if index >= (self.k-1):
                index=self.k-1
            helper = np.array([traj['x'][i], traj['y'][i]])
            LogProb = LogProb + cluster.logValueOfGaussuan2D(self.mean[index], self.variance[index], helper)


        return LogProb

    @staticmethod
    def findLowestIndex(clusters):
        index = 0
        MinProb = 0
        for i, c in enumerate(clusters):
            l = len(c.trajectories)
            prob = 0
            if l == 0:
                return i
            for p in c.trajectoriesProb:
                prob = prob + p
            prob = prob / l
            if MinProb > prob:
                MinProb = prob
                index = i
        return index

    @staticmethod
    def findLEastLiklyTraj(clusters):
        index = 0
        MinProb = 0
        for c in clusters:
            for i, p in zip(c.trajectoriesIndex, c.trajectoriesProb):
                if p < MinProb:
                    MinProb = p
                    index = i
        return index


    @staticmethod
    def Expectation(d, trajectories):
        labels=[]
        clusters = copy.deepcopy(d)
        for c in clusters:
            c.trajectories = []
            c.trajectoriesProb = []
            c.trajectoriesIndex = []
        for j, traj in enumerate(trajectories):
            maxprob = 0
            index = 0
            for i, c in enumerate(clusters):
                if d[i].trajectoriesIndex == []:
                    continue
                prob = c.probabilityOfTrajInCluster(traj)
                if i == 0:
                    maxprob = prob
                    continue
                if maxprob < prob:
                    maxprob = prob
                    index = i
            clusters[index].trajectories.append(traj)
            clusters[index].trajectoriesProb.append(maxprob)
            clusters[index].trajectoriesIndex.append(j)
            labels.append(index)
        return clusters,labels
    @staticmethod
    def ExpectationOutliers(d, trajectories,labels):
        clusters = copy.deepcopy(d)

        for j, traj in enumerate(trajectories):
            maxprob = 0
            index = 0
            for i, c in enumerate(clusters):
                if labels!= -1:
                    continue
                prob = c.probabilityOfTrajInCluster(traj)
                if i == 0:
                    maxprob = prob
                    continue
                if maxprob < prob:
                    maxprob = prob
                    index = i
            clusters[index].trajectories.append(traj)
            clusters[index].trajectoriesProb.append(maxprob)
            clusters[index].trajectoriesIndex.append(j)
        return clusters

    def Maximization(self):
        for i in range(self.k):
            points = []
            for traj in self.trajectories:
                for b in range(self.beta * i, self.beta * (i + 1)):
                    points.append([traj.to_numpy()[b][0], traj.to_numpy()[b][1]])
            self.mean[i] = np.mean(points, axis=0)
            self.variance[i] = np.cov(np.array(points).transpose())

            #print(points)
            #print(self.variance )
            #print(self.mean)
            #print(len(self.trajectories))
            #print(self.beta)
            #input()


    @staticmethod
    def logValueOfGaussuan2D(mean, variance, value):
        # print(value)
        # print(mean)
        dif = value - mean
        # print(variance)
        return -np.log(2 * math.pi) - 0.5 * np.log(np.linalg.det(variance)) - 0.5 * np.matmul(
            np.matmul(dif.transpose(), np.linalg.inv(variance)), dif)

    @staticmethod
    def BIC(clusters):
        k= 0
        n=0
        logLiklyhood=0
        for c in clusters:
            k=k+c.k*6
            n=n+len(c.trajectories)*200
            for traj in c.trajectories:
                logLiklyhood=logLiklyhood+c.probabilityOfTrajInCluster(traj)
        n=np.log(n)
        return k*n-2*logLiklyhood

    @staticmethod
    def AIC(clusters):
        k= 0
        logLiklyhood=0
        for c in clusters:
            k=k+c.k*6
            for traj in c.trajectories:
                logLiklyhood=logLiklyhood+c.probabilityOfTrajInCluster(traj)
        return 2*k-2*logLiklyhood