/
plotting.py
executable file
·224 lines (210 loc) · 8.5 KB
/
plotting.py
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
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
from __future__ import division
from sys import platform as _platform
import matplotlib
if _platform == "linux" or _platform == "linux2":
matplotlib.use('Agg')
from graph_tool.all import *
import matplotlib.pylab as plt
import matplotlib.cm as colormap
#import seaborn
import pandas as pd
import utils
import numpy as np
from graph_tool.all import *
import scipy.stats as stats
import utils
import datetime
import os
def create_scatter(x, y, fname, **kwargs):
matplotlib.rcParams.update({'font.size': 15})
assert isinstance(x, tuple)
assert isinstance(y, tuple)
x_label, x_data = x
y_label, y_data = y
x_data = np.array(x_data)
y_data = np.array(y_data)
y_data, y_data_mod = utils.shift_data_pos(y_data)
x_data, x_data_mod = utils.shift_data_pos(x_data)
#df = pd.DataFrame(columns=[x_label], data=x_data)
#df[y_label] = y_data
alpha = 1 / np.log10(len(y_data))
f, ax = plt.subplots()
x_data_log, y_data_log = np.log10(x_data), np.log10(y_data)
logarithmic_pearson = stats.pearsonr(x_data_log, y_data_log)[0]
pearson = stats.pearsonr(x_data, y_data)[0]
if logarithmic_pearson > .2 or logarithmic_pearson < -.2:
coefs = np.polyfit(x_data_log, y_data_log, deg=1)
ax.plot(None, lw=0, c='white', alpha=0., label='k: ' + "%.2f" % coefs[1])
else:
coefs = None
if not np.isnan(logarithmic_pearson):
ax.plot(None, lw=0, c='white', alpha=0., label='log10 pearson: ' + "%.2f" % logarithmic_pearson)
ax.plot(None, lw=0, c='white', alpha=0., label='pearson: ' + "%.2f" % pearson)
for i in range(3):
if i == 0:
filt = y_data > 1
label = 'increased'
marker = '^'
c = 'red'
elif i == 1:
filt = np.isclose(y_data, 1.)
label = 'neutral'
c = 'gray'
marker = 'o'
else:
filt = y_data < 1
label = 'decreased'
c = 'blue'
marker = 'v'
x_filt, y_filt = x_data[filt], y_data[filt]
ax.scatter(x=x_filt, y=y_filt, alpha=alpha, s=70, color=c, lw=0, label=label, marker=marker, **kwargs)
plt.axhline(1., color='red', alpha=.25, lw=2, ls='--')
y_min, y_max = y_data.min(), y_data.max()
x_min, x_max = x_data.min(), x_data.max()
if coefs is not None:
lin_space = np.linspace(np.log10(x_min), np.log10(x_max), 100)
y_log_space = (coefs[1] + lin_space * coefs[0])
ax.plot(10 ** lin_space, 10 ** y_log_space, lw=4, alpha=0.9, label='logarithmic fit', c='green')
ax.set_xlim([x_min, x_max])
ax.set_ylim([y_min, y_max])
if not np.isnan(logarithmic_pearson):
if np.isclose(logarithmic_pearson, 0.):
loc = 'best'
elif logarithmic_pearson > 0:
loc = 'upper left'
elif logarithmic_pearson < 0:
loc = 'upper right'
else:
loc = 'best'
else:
loc = 'best'
loc = 'best'
plt.legend(loc=loc)
plt.xlabel(x_label + (' (shifted)' if x_data_mod else ''))
plt.ylabel(y_label + (' (shifted)' if y_data_mod else ''))
ax.set_xscale('log')
ax.set_yscale('log')
plt.tight_layout()
plt.savefig(fname)
plt.close('all')
def draw_graph(network, color, min_color=None, max_color=None, groups=None, sizep=None, colormap_name='bwr', min_vertex_size_shrinking_factor=4, output='graph.png', output_size=(15, 15), dpi=80, standardize=False, color_bar=True, crop=True, **kwargs):
output_splitted = output.rsplit('/', 1)[-1].split('_graph_')
net_name, prop_key = output_splitted[0], output_splitted[-1]
print_prefix = utils.color_string('[' + net_name + '] ') + '[' + prop_key + '] [' + str(
datetime.datetime.now().replace(microsecond=0)) + '] draw graph'
print print_prefix
print_prefix += ': '
num_nodes = network.num_vertices()
min_vertex_size_shrinking_factor = min_vertex_size_shrinking_factor
if num_nodes < 10:
num_nodes = 10
max_vertex_size = np.sqrt((np.pi * (min(output_size) * dpi / 2) ** 2) / num_nodes)
if max_vertex_size < min_vertex_size_shrinking_factor:
max_vertex_size = min_vertex_size_shrinking_factor
min_vertex_size = max_vertex_size / min_vertex_size_shrinking_factor
if sizep is None:
sizep = max_vertex_size + min_vertex_size
sizep /= 3
else:
sizep = prop_to_size(sizep, mi=min_vertex_size / 3 * 2, ma=max_vertex_size / 3 * 2, power=2)
v_shape = 'circle'
if isinstance(groups, str):
try:
v_shape = network.vp[groups].copy()
#groups = network.vp[groups]
#unique_groups = set(np.array(groups.a))
#num_groups = len(unique_groups)
#groups_c_map = colormap.get_cmap('gist_rainbow')
#groups_c_map = {i: groups_c_map(idx / (num_groups - 1)) for idx, i in enumerate(unique_groups)}
#v_pen_color = network.new_vertex_property('vector<float>')
#for v in network.vertices():
# v_pen_color = groups_c_map[groups[v]]
v_shape.a %= 14
except KeyError:
# print print_prefix + 'cannot find groups property:', groups
v_shape = 'circle'
cmap = colormap.get_cmap(colormap_name)
color = color.copy()
v_shape = network.new_vertex_property('int')
v_shape.a = np.array(
[0 if np.isclose(color[int(v)], 1.) else (1 if color[int(v)] > 1. else 4) for v in network.vertices()],
dtype='int')
try:
_ = color.a
except AttributeError:
c = network.new_vertex_property('float')
c.a = color
color = c
min_color = color.a.min() if min_color is None else min_color
max_color = color.a.max() if max_color is None else max_color
if np.isclose(min_color, max_color):
min_color = 0
max_color = 2
#orig_color = np.array(color.a)
if standardize:
color.a -= color.a.mean()
color.a /= color.a.var()
color.a += 1
color.a /= 2
else:
#color.a -= min_color
#color.a /= max_color
tmp = np.array(color.a)
tmp[tmp > 1] = 1 + (tmp[tmp > 1] / (max_color/1))
color.a = tmp
color.a /= 2
if not output.endswith('.png'):
output += '.png'
color_pmap = network.new_vertex_property('vector<float>')
tmp = np.array([np.array(cmap(i)) for i in color.a])
color_pmap.set_2d_array(tmp.T)
plt.switch_backend('cairo')
f, ax = plt.subplots(figsize=(15, 15))
output_size = (output_size[0], output_size[1]*.3) # make space for colorbar
edge_alpha = 0.3 if network.num_vertices() < 1000 else 0.01
pen_width = 0.8 if network.num_vertices() < 1000 else 0.1
v_pen_color = [0., 0., 0., 1] if network.num_vertices() < 1000 else [0.0, 0.0, 0.0, edge_alpha]
graph_draw(network, vertex_fill_color=color_pmap, mplfig=ax, vertex_pen_width=pen_width, vertex_shape=v_shape,
vertex_color=v_pen_color, edge_color=[0.179, 0.203, 0.210, edge_alpha], vertex_size=sizep,
output_size=output_size, output=output, **kwargs)
if color_bar:
cmap = plt.cm.ScalarMappable(cmap=cmap)
cmap.set_array([0., 2.])
cbar = f.colorbar(cmap, drawedges=False)
ticks = [0, 1.0, max_color / 1]
cbar.set_ticks([0., 1., 2.])
tick_labels = None
non_zero_dig = 1
for digi in range(10):
tick_labels = [str("{:2." + str(digi) + "f}").format(i) for i in ticks]
if any([len(i.replace('.', '').replace('0', '').replace(' ', '').replace('-', '').replace('+', '')) > 0 for
i in tick_labels]):
non_zero_dig -= 1
if non_zero_dig == 0:
break
cbar.ax.set_yticklabels(tick_labels)
cbar.ax.tick_params(labelsize=40)
#var = stats.tvar(orig_color)
cbar.set_label('Bias Factor', labelpad=+30)
matplotlib.rcParams.update({'font.size': 40})
plt.axis('off')
plt.savefig(output, bbox_tight=True, dpi=dpi)
plt.close('all')
plt.switch_backend('Agg')
if crop:
if output.endswith('.pdf'):
os.system('pdfcrop ' + output + ' ' + output)
else:
pass
#os.system('convert ' + output + ' -trim ' + output)
# print print_prefix + 'done'
def plot_stat_dist(ser, output_filename, **kwargs):
matplotlib.rcParams.update({'font.size': 15})
assert isinstance(ser, pd.Series)
ser.plot(kind='hist', **kwargs)
# plt.title(key)
plt.ylabel('#nodes')
plt.xlabel('stationary value')
plt.tight_layout()
plt.savefig(output_filename)
plt.close('all')