CLoNe/plot.py at master · LBM-EPFL/CLoNe · 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
import numpy as np
from itertools import cycle, islice
from scipy.stats.kde import gaussian_kde
from sklearn.decomposition import PCA

import statsmodels.api as sm
import matplotlib.pyplot as mplot
import matplotlib.cm as cm
from mpl_toolkits.mplot3d import Axes3D


def plot_clusters(clone, data, path, headers=None):
    color_list = np.array(['yellowgreen', 'orange',  'crimson', 'mediumpurple', 'deepskyblue', 'Aquamarine', 'DarkGoldenRod',
              'Khaki', 'SteelBlue', 'Olive', 'Violet', 'DarkSeaGreen', 'RosyBrown', 'LightPink', 'DodgerBlue',
              'lightcoral', 'chocolate', 'burlywood', 'cyan', 'olivedrab', 'palegreen', 'turquoise', 'gold', 'teal',
              'hotpink', 'moccasin', 'lawngreen', 'sandybrown', 'blueviolet', 'powderblue', 'plum', 'springgreen',
              'mediumaquamarine', 'rebeccapurple', 'peru', 'lightsalmon', 'khaki', 'sienna', 'lightseagreen', 'lightcyan'])
    colors = np.array(list(islice(cycle(color_list),len(clone.centers))))

    if headers is None:
        headers = ["C%i"%x for x in range(data.shape[1])]
    if data.shape[1] == 1:
        plot_1d_clusters(clone, data, path, headers, colors)
    elif data.shape[1] == 2:
        plot_2d_clusters(clone, data, path, headers, colors)
    else:
        plot_nd_clusters(clone, data, path, headers, colors)
    mplot.show()


def plot_1d_clusters(clone, data, path, headers, colors):
    centers = np.array(clone.centers)
    labels = np.array(clone.labels_)
    labels_all = np.array(clone.labels_all)
    core = clone.core_card
    rho = clone.rho

    # Mask for plotting
    assigned_mask = np.where(labels != -1)
    outliers_mask = np.where(labels == -1)

    # KDE
    kde = sm.nonparametric.KDEUnivariate(data.astype(np.float))
    kde.fit()

    # Sort some values for better visualization after
    arcore = np.argsort(core)
    s_cores = core[arcore]
    s_x = data[arcore]

    # Plot core
    mplot.figure(figsize=(4, 2))
    mplot.scatter(s_x, [0] * len(data), marker='|', linewidth=0.1, s=150, c=s_cores, cmap=cm.nipy_spectral)
    mplot.yticks([])
    mplot.xlabel(headers[0], fontsize=15)
    cbar = mplot.colorbar()
    mplot.tick_params(axis='x', which='major', length=8, width=2, labelsize=15)
    cbar.ax.tick_params(axis='x', which='major', length=8, width=2, labelsize=15)
    cbar.ax.set_xlabel("#core", fontsize=15)
    mplot.tight_layout()
    mplot.savefig(path + "/cores.png", dpi=300)

    # Plot clusters
    mplot.figure(figsize=(4, 4))
    mplot.scatter(data[assigned_mask, 0], [0] * len(data[assigned_mask]), marker='|', color=colors[labels[assigned_mask]])
    mplot.scatter(data[outliers_mask, 0], [0] * len(data[outliers_mask]), marker='x', s=10, color='black')
    mplot.scatter(data[centers, 0], [0] * len(data[centers]), marker='*', s=100, color='black')
    mplot.plot(kde.support, kde.density, c="grey", marker="None", linestyle="--")
    mplot.xlabel(headers[0])
    mplot.ylabel("KDE")

    mplot.tight_layout()
    mplot.savefig(path + "/clusters.png", dpi=300)


def plot_2d_clusters(clone, data, path, headers, colors):
    centers = np.array(clone.centers)
    labels = np.array(clone.labels_)
    core_card = clone.core_card
    rho = clone.rho

    # Core cardinality mapped to each point
    size = (3.5,4)
    cluster_fig = mplot.figure(figsize=size)

    arcore = np.argsort(core_card)
    s_cores = core_card[arcore]
    s_data = data[arcore]

    mplot.scatter(s_data[:, 0], s_data[:, 1], marker='o', c=s_cores, cmap=cm.nipy_spectral)
    mplot.xlabel("#core", labelpad=20, fontsize=15)
    cbar = mplot.colorbar(orientation='horizontal')
    cbar.ax.tick_params(axis='x', which='major', length=8, width=2, labelsize=15)
    cbar.ax.tick_params(axis='y', which='major', length=8, width=2, labelsize=15)
    mplot.scatter(data[centers, 0], data[centers, 1], marker='*', s=150, c='black', cmap=cm.nipy_spectral)
    mplot.xticks([])
    mplot.yticks([])
    mplot.tight_layout()
    mplot.savefig(path + "/cores.png")

    # Cluster plot by labels
    size = (4,4)
    cluster_fig = mplot.figure(figsize=size)
    ax_clus = cluster_fig.add_subplot(111)
    assigned_mask = np.where(labels != -1)
    outliers_mask = np.where(labels == -1)
    ax_clus.set_xlabel(headers[0])
    ax_clus.set_ylabel(headers[1])
    ax_clus.scatter(data[outliers_mask, 0], data[outliers_mask, 1], marker='x', s=10, color='black', alpha=0.7)
    ax_clus.scatter(data[assigned_mask, 0], data[assigned_mask, 1], marker='o', s=30, edgecolors='black', linewidth=0.1, color=colors[labels[assigned_mask]])
    ax_clus.scatter(data[centers, 0], data[centers, 1], marker='*', s=150, c='black', cmap=cm.nipy_spectral)
    mplot.tight_layout()
    mplot.savefig(path + "/clusters.png")


def plot_nd_clusters(clone, data, path, headers, colors):
    centers = np.array(clone.centers)
    nb_clust = len(centers)
    data_dim = data.shape[1]
    labels = np.array(clone.labels_)
    labels_all = np.array(clone.labels_all)
    core = clone.core_card
    rho = clone.rho

    # Mask for plotting
    assigned_mask = np.where(labels != -1)
    outliers_mask = np.where(labels == -1)

    fig_3d = mplot.figure(figsize=(4,4))
    ax = fig_3d.add_subplot(111, projection='3d')
    ax.scatter(data[outliers_mask, 0], data[outliers_mask, 1], data[outliers_mask, 2], marker='.', color='black')
    ax.scatter(data[assigned_mask, 0], data[assigned_mask, 1], data[assigned_mask, 2], marker='o', color=colors[labels[assigned_mask]])
    ax.set_xlabel(headers[0])
    ax.set_ylabel(headers[1])
    ax.set_zlabel(headers[2])
    mplot.savefig(path + "/clusters_3D.png", dpi=300)

    # Make plot
    fig = mplot.figure(figsize=(4,4))
    for n in range(data_dim):
        ax = fig.add_subplot(data_dim, 1, n+1)
        ax.set_xlabel("")

        # Plot distributions per cluster
        for clus_idx in range(len(centers)):
            cur_coords = np.array(data[labels == clus_idx, n],ndmin=2)
            p = ax.violinplot(cur_coords[0], [clus_idx+1], showmeans=False, showextrema=False, showmedians=False, widths=0.8)
            for pc in p['bodies']:
                pc.set_facecolor(colors[clus_idx])
                pc.set_edgecolor('black')
                pc.set_alpha(1)

        mplot.yticks(rotation=90, va="center")
        ax.set_ylabel(headers[n], labelpad=10, rotation=90)

        ax.grid(True, which='major', axis='y', linestyle='--', c='black')

        # Set x axis only for last dimension
        ax.set_xticks(np.arange(0,nb_clust+2, 1))
        ax.get_xaxis().set_visible(False)
        if n == data_dim - 1:
            ax.get_xaxis().set_visible(True)
            labels = []
            for n in range(nb_clust):
                labels.append("C%i"%(n+1))
            ax.set_xticks(np.arange(1,nb_clust+2, 1))
            ax.set_xticklabels(labels)
            mplot.tick_params(axis='x', pad=20)
        else:
            ax.get_xaxis().set_visible(False)
        ax.set_xlim(xmin=0.3, xmax=(0.8+nb_clust))
    mplot.xticks(rotation=90, va="center")
    mplot.tight_layout()
    mplot.savefig(path + "/clusters_violin.png", dpi=300)