babusca/graph.py at master · georglind/babusca · 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
156
157
158
159
160
161
162
163
164
from __future__ import print_function, division
import numpy as np
import matplotlib.pyplot as plt


class Graph:
    def __init__(self, nodes, xy=None, links=None):
        self.N = len(nodes)
        self.nodes = nodes

        if xy is None:
            xy = np.zeros((self.N, 2))
        self.xy = xy

        if links is None:
            links = []

        self.links = links

        neighbors = []
        for i in xrange(self.N):
            n = [j[0] for j in links if j[1] == i] + [j[1] for j in links if j[0] == i]
            neighbors.append(n)
        self.neighbors = neighbors

    def forcedirectedlayout(self):
        converged, self.xy = FD(self)

    def plot(self, fig=None, ax=None):
        xy = self.xy[:, :]

        xymax = np.max(xy, axis=0)
        xymin = np.min(xy, axis=0)

        # imin = [np.where(xy[:, i] == xymin[i])[0] for i in xrange[2]]
        xy -= np.reshape(np.tile(.5 * (xymax + xymin), self.N), (self.N, 2))

        xyscale = np.max((xymax - xymin) / 100)

        show = False
        if fig is None or ax is None:
            show = True
            fig, ax = plt.subplots(figsize=(4, 4))

        for link in self.links:
            ax.plot(xy[link[0:2], 0] / xyscale, xy[link[0:2], 1] / xyscale, 'b')

        for i in xrange(self.N):
            x, y = xy[i, 0:2] / xyscale
            circle = plt.Circle((x, y), 3, color='#e5e5e5', zorder=10)
            ax.add_artist(circle)
            ax.text(x, y + 1, '{0}'.format(i), zorder=20, ha='center', va='center')

        ax.set_xlim([-60, 60])
        ax.set_ylim([-60, 60])

        if show is True:
            plt.show()


def FD(graph):
    """ Force directed layout routine """
    energy = 0
    heat = 0.02

    xy = 1 * np.random.rand(graph.N, 2)
    vs = np.zeros((graph.N, 2))

    settings = {'vmax': 1, 'friction': .2}

    cnt = 0
    converged = False
    while not converged:
        cnt += 1

        if heat > 0:
            if cnt == 20:
                heat = .01
            elif cnt == 40:
                heat = .005
            elif cnt == 80:
                heat = 0

        xy, vs, energy = FDstep(graph, xy, vs, energy, heat, settings)

        if energy > 100 * graph.N:
            break

        # print(np.sum(np.abs(vs))/graph.N)
        converged = np.sum(np.abs(vs)) / graph.N < 1e-3
        if cnt > 500:
            converged = True

    print(cnt)
    return converged, xy


def FDstep(graph, xy=None, vs=None, energy=0, heat=0.5, settings=None):
    """ Force directed layout step"""
    N = graph.N

    if xy is None:
        xy = np.random.rand(N, 2)
    if vs is None:
        vs = np.zeros((N, 2))

    # xy += -np.reshape(np.tile(np.sum(xy, axis=0)/N, N), (N, 2))

    nenergy = 0
    rforce = np.zeros((N, 2))  # repelling force
    aforce = np.zeros((N, 2))  # attractive force

    for i in xrange(N):
        for j in xrange(N):
            if i == j:
                continue
            d = xy[i, :] - xy[j, :]
            r = np.sqrt(np.sum(d ** 2))
            rforce[i, :] += 0.001 * d / r ** 3 if r > 0.001 else .1 * (np.random.rand(2) - .5)
            nenergy += 0.005 / max(r, 0.001)  # new energy

    for i in xrange(N):
        for j in graph.neighbors[i]:
            d = xy[i, :] - xy[j, :]
            r = np.sqrt(np.sum(d ** 2))
            aforce[i, :] += -2 * d / 100 if r > 0.001 else .1 * (np.random.rand(2) - .5)
            nenergy += r ** 2 / 100

    vs += (aforce + rforce)

    # idx = np.where(np.abs(vs) > settings['vmax'])
    # vs[idx] = settings['vmax']*np.sign(vs[idx])

    vs -= settings['friction'] * vs
    # idx = np.where(np.abs(vs) > settings['friction'])
    # vs[idx] += -settings['friction']*np.sign(vs[idx])

    if heat > 0:
        vs += 2 * heat * (2 * np.random.rand(N, 2) - 1)

    xy += vs
    # if heat > 0:
    #     xy +=

    xy += -np.reshape(np.tile(np.sum(xy, axis=0), N) / N, (N, 2))

    # print(vs)
    # graph.xy = xy
    # graph.plot()

    return xy, vs, nenergy


if __name__ == "__main__":
    # This is a small demonstration of this graph library
    # showing how to plot of a 7 site chain using
    # force-directed layout.

    nodes = [0] * 7
    links = [[0, 1], [1, 2], [2, 3], [3, 4], [4, 0], [4, 6], [6, 5]]

    g = Graph(nodes, None, links)
    g.forcedirectedlayout()
    g.plot()