ad_examples/python/aad/plot_anomalies_rectangle.py at master · Wolgo/ad_examples · 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
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
import os
import numpy as np
import numpy.random as rnd
import matplotlib.pyplot as plt

import logging
from pandas import DataFrame

from common.gen_samples import *
from common.data_plotter import *

from aad.aad_globals import *
from aad.aad_support import *
from aad.forest_description import *
from aad.anomaly_dataset_support import *
# from percept.percept import *


"""
pythonw -m aad.plot_anomalies_rectangle
"""


def get_x_tau(x, w, tau):
    v = x.dot(w)
    ranked = np.argsort(-v)
    tau_id = ranked[int(tau * len(v))]
    return tau_id, x[tau_id]


def plot_anomalies_ifor(outdir, plot=False, plot_legends=False):
    u_theta = np.pi * 4. / 4 + np.pi * 5 / 180
    x, y = get_sphere_samples([(50, 0, np.pi * 4. / 4, np.pi * 4. / 4 + np.pi * 2 / 4),
                               (15, 1, u_theta - np.pi * 5 / 180, u_theta + np.pi * 5 / 180),
                               (15, 1, np.pi * 6. / 4 - np.pi * 1.5 / 180, np.pi * 6. / 4)])
    n, d = x.shape

    id_nomls = np.where(y == 0)[0]
    id_anoms = np.where(y == 1)[0]
    n_anoms = len(id_anoms)

    x_nomls, y_nomls = x[id_nomls, :], y[id_nomls]
    x_anoms, y_anoms = x[id_anoms, :], y[id_anoms]

    if plot:
        axis_fontsize = 16

        line_colors = ["blue", "red", "red"]
        line_types = ["--", "--", "-"]
        line_widths = [2, 2, 2]
        lines = list()
        line_labels = list()

        tau = n_anoms * 1. / n  # multiplying by a factor to move the plane lower
        w = normalize(np.ones(2))
        r = np.array([np.min(x[:, 0]), np.max(x[:, 0])])

        tau_id, x_tau = get_x_tau(x, w, tau)
        q_tau = w.dot(x_tau)

        # plot the true weight vector
        u = interpolate_2D_line_by_point_and_vec(np.array([-1., 1.]), [0., 0.],
                                                 [np.cos(u_theta + np.pi * 1 / 4), np.sin(u_theta + np.pi * 1 / 4)])
        lines.append(u)
        line_labels.append(r"True weights ${\bf u}$")

        zd = interpolate_2D_line_by_point_and_vec(np.array([-1., 1.0]), [0., 0.], w)
        lines.append(zd)
        line_labels.append(r"Uniform weights ${\bf w}_{unif}$")

        zw = interpolate_2D_line_by_slope_and_intercept(np.array([-1., 1.]), -w[0] / w[1], q_tau / w[1])
        lines.append(zw)
        line_labels.append(r"hyperplane $\perp$ ${\bf w}_{unif}$")

        pdffile = os.path.join(outdir, "anomalies_in_ifor.pdf")
        dp = DataPlotter(pdfpath=pdffile, rows=1, cols=1)
        pl = dp.get_next_plot()
        pl.set_aspect('equal')
        # plt.xlabel('x', fontsize=axis_fontsize)
        # plt.ylabel('y', fontsize=axis_fontsize)
        plt.xticks([])
        plt.yticks([])
        plt.xlim([-1.05, 1.05])
        plt.ylim([-1.05, 1.05])
        pl.scatter(x_nomls[:, 0], x_nomls[:, 1], s=45, c="blue", marker="+", label="Nominal")
        pl.scatter(x_anoms[:, 0], x_anoms[:, 1], s=45, c="red", marker="+", label="Anomaly")
        for i, line in enumerate(lines):
            color = "blue" if line_colors is None else line_colors[i]
            pl.plot(line[:, 0], line[:, 1], line_types[i], color=color, linewidth=line_widths[i],
                    label=line_labels[i] if plot_legends else None)

        plt.axhline(0, linestyle="--", color="lightgrey")
        plt.axvline(0, linestyle="--", color="lightgrey")

        if plot_legends:
            pl.legend(loc='lower right', prop={'size': 12})
        dp.close()
    return x, y


def plot_anomalies_rect(outdir, plot=False, plot_legends=False):
    x_nomls = rnd.uniform(0., 1., 500)
    x_nomls = np.reshape(x_nomls, newshape=(250, -1))

    anom_mu = (0.83, 0.95)
    u_theta = np.arctan(0.9 / 0.8)

    anom_score_dist = MVNParams(
        mu=np.array([anom_mu[0], anom_mu[1]]),
        mcorr=np.array([
            [1, -0.5],
            [0, 1.0]]),
        dvar=np.array([0.002, 0.0005])
    )
    n_anoms = 30
    x_anoms = generate_dependent_normal_samples(n_anoms,
                                                anom_score_dist.mu,
                                                anom_score_dist.mcorr,
                                                anom_score_dist.dvar)

    x = np.vstack([x_nomls, x_anoms])
    y = np.array(np.zeros(x_nomls.shape[0], dtype=int))
    y = np.append(y, np.ones(x_anoms.shape[0], dtype=int))

    if plot:
        n, d = x.shape

        # tau is computed assuming that the anomalies occupy tau-proportion
        # of the circumference
        tau = n_anoms * 1.3 / n  # multiplying by a factor to move the plane lower
        w = normalize(np.ones(2))
        r = np.array([np.min(x[:, 0]), np.max(x[:, 0])])

        line_colors = ["blue", "red", "red"]
        line_types = ["--", "--", "-"]
        line_widths = [2, 2, 2]
        lines = list()
        line_labels = list()

        tau_id, x_tau = get_x_tau(x, w, tau)
        q_tau = w.dot(x_tau)

        # plot the true weight vector
        u = interpolate_2D_line_by_point_and_vec(np.array([0., 1.]), [0., 0.],
                                                 [np.cos(u_theta), np.sin(u_theta)])
        lines.append(u)
        line_labels.append(r"True weights ${\bf u}$")

        zd = interpolate_2D_line_by_point_and_vec(np.array([0., 1.0]), [0., 0.], w)
        lines.append(zd)
        line_labels.append(r"Uniform weights ${\bf w}_{unif}$")

        zw = interpolate_2D_line_by_slope_and_intercept(np.array([0., 1.05]), -w[0] / w[1], q_tau / w[1])
        lines.append(zw)
        line_labels.append(r"hyperplane $\perp$ ${\bf w}_{unif}$")

        axis_fontsize = 16
        pdffile = os.path.join(outdir, "anomalies_in_rect.pdf")
        dp = DataPlotter(pdfpath=pdffile, rows=1, cols=1)
        pl = dp.get_next_plot()
        pl.set_aspect('equal')
        # plt.xlabel('x', fontsize=axis_fontsize)
        # plt.ylabel('y', fontsize=axis_fontsize)
        plt.xticks([])
        plt.yticks([])
        plt.xlim([-0.05, 1.05])
        plt.ylim([-0.05, 1.05])
        pl.scatter(x_nomls[:, 0], x_nomls[:, 1], s=45, c="blue", marker="+", label="Nominal")
        pl.scatter(x_anoms[:, 0], x_anoms[:, 1], s=45, c="red", marker="+", label="Anomaly")
        for i, line in enumerate(lines):
            color = "blue" if line_colors is None else line_colors[i]
            pl.plot(line[:, 0], line[:, 1], line_types[i], color=color, linewidth=line_widths[i],
                    label=line_labels[i] if plot_legends else None)

        if plot_legends:
            pl.legend(loc='lower right', prop={'size': 12})
        dp.close()

    return x, y


if __name__ == "__main__":

    logger = logging.getLogger(__name__)

    args = get_command_args(debug=True, debug_args=["--debug",
                                                    "--plot",
                                                    "--log_file=temp/plot_anomalies_rectangle.log"])
    # print "log file: %s" % args.log_file
    configure_logger(args)

    rnd.seed(42)

    outdir = "./temp/illustration"
    dir_create(outdir)

    # plot isolation forest score distribution illustration
    # plot_anomalies_ifor(outdir, plot=True, plot_legends=False)

    plot_anomalies_rect(outdir, plot=True, plot_legends=False)