File size: 4,576 Bytes
e11c92f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""Test diffusion."""

import matplotlib.pyplot as plt
import numpy as np
from numpy.testing import assert_array_almost_equal

from MARBLE import construct_dataset
from MARBLE import dynamics
from MARBLE import geometry
from MARBLE import plotting
from MARBLE.layers import Diffusion


def f1(x):
    """f1"""
    eps = 1e-1
    norm = np.sqrt((x[:, [0]] - 1) ** 2 + x[:, [1]] ** 2 + eps)
    u = x[:, [1]] / norm
    v = -(x[:, [0]] - 1) / norm
    return np.hstack([u, v])


def f2(x):
    """f2"""
    y = []
    for _ in range(x.shape[0]):
        y_ = np.random.uniform(size=(3))
        y_ /= np.linalg.norm(y_)
        y.append(y_)

    return np.vstack(y)


def sphere():
    """sphere"""
    u, v = np.mgrid[0 : 2 * np.pi : 20j, 0 : np.pi : 11j]
    x = np.cos(u) * np.sin(v)
    y = np.sin(u) * np.sin(v)
    z = np.cos(v)
    return np.vstack([x.flatten(), y.flatten(), z.flatten()]).T


def test_diffusion(plot=False):
    """Test diffusion and laplacian creation."""
    # parameters
    n = 512
    k = 30
    tau0 = 50

    # f1: constant, f2: linear, f3: parabola, f4: saddle
    x = dynamics.sample_2d(n, [[-1, -1], [1, 1]], "random")
    y = f1(x)  # evaluated functions

    # #construct PyG data object
    data = construct_dataset(x, y, graph_type="cknn", k=k)

    gauges, _ = geometry.compute_gauges(data)
    assert_array_almost_equal(
        gauges.numpy()[:5],
        np.array(
            [
                [[-0.19064367, 0.9816593], [0.9816593, 0.19064367]],
                [[-0.97356814, 0.22839674], [0.22839674, 0.97356814]],
                [[-0.91470975, -0.40411147], [-0.40411147, 0.91470975]],
                [[-0.1206701, 0.99269265], [0.99269265, 0.1206701]],
                [[-0.37872583, 0.9255089], [0.9255089, 0.37872583]],
            ]
        ),
        decimal=5,
    )
    R = geometry.compute_connections(data, gauges)
    assert_array_almost_equal(
        R.to_dense()[:5, :5],
        np.array(
            [
                [1.0, 0.0, 0.0, 0.0, -0.22231616],
                [0.0, 1.0, 0.0, 0.0, -0.97497463],
                [0.0, 0.0, 1.0, 0.0, 0.0],
                [0.0, 0.0, 0.0, 1.0, 0.0],
                [-0.22231616, -0.97497463, 0.0, 0.0, 1.0],
            ],
        ),
        decimal=5,
    )
    L = geometry.compute_laplacian(data)
    assert_array_almost_equal(
        L.to_dense().numpy()[:5, :5],
        np.array(
            [
                [1.0, 0.0, -0.01967779, 0.0, 0.0],
                [0.0, 1.0, 0.0, 0.0, 0.0],
                [-0.02420571, 0.0, 1.0, 0.0, 0.0],
                [0.0, 0.0, 0.0, 1.0, 0.0],
                [0.0, 0.0, 0.0, 0.0, 1.0],
            ],
        ),
        decimal=5,
    )
    Lc = geometry.compute_connection_laplacian(data, R)
    assert_array_almost_equal(
        Lc.to_dense().numpy()[:5, :5],
        np.array(
            [
                [1.0, 0.0, 0.0, 0.0, 0.00437469],
                [0.0, 1.0, 0.0, 0.0, 0.01918535],
                [0.0, 0.0, 1.0, 0.0, 0.0],
                [0.0, 0.0, 0.0, 1.0, 0.0],
                [0.00538132, 0.02359995, 0.0, 0.0, 1.0],
            ],
        ),
        decimal=5,
    )

    diffusion = Diffusion(tau0=tau0)
    data.x = diffusion(data.x, L, Lc, method="matrix_exp")
    assert_array_almost_equal(
        data.x.detach().numpy()[:5],
        np.array(
            [
                [0.8945822, 0.2413084],
                [-0.06356261, 0.9169159],
                [-0.5462601, 0.7629167],
                [0.9424986, -0.01397797],
                [0.57801795, 0.68634313],
            ]
        ),
        decimal=5,
    )

    if plot:
        plotting.fields(data)
        plt.show()


def test_diffusion_sphere(plot=False):
    """Test diffusion on sphere."""
    # parameters
    k = 0.4
    tau0 = 10.0

    x = sphere()
    y = f2(x)

    # construct PyG data object
    data = construct_dataset(
        x, y, graph_type="radius", k=k, frac_geodesic_nb=1.5, var_explained=0.9
    )

    L = geometry.compute_laplacian(data)

    diffusion = Diffusion(tau0=tau0)
    data.x = diffusion(data.x, L, method="matrix_exp")
    assert_array_almost_equal(
        data.x.detach().numpy()[:5],
        np.array(
            [
                [0.513162, 0.44882008, 0.5685046],
                [0.35709542, 0.67346, 0.44372997],
                [0.32471117, 0.3551194, 0.81424344],
                [0.6844833, 0.53020036, 0.4575338],
                [0.5897326, 0.68115395, 0.41908088],
            ]
        ),
        decimal=5,
    )

    if plot:
        plotting.fields(data, alpha=1)
        plt.show()