-
Notifications
You must be signed in to change notification settings - Fork 2
/
plot_gap_distribution.py
177 lines (147 loc) · 5.85 KB
/
plot_gap_distribution.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
import matplotlib as mpl
mpl.use('Agg')
from fractions import Fraction
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import FixedLocator
import platform
import sys
import math
import statistics
try:
import cPickle as pickle
except ImportError: # Python 3.x
import pickle
test_names = []
method_names = []
colors = ["#F78154", "#1C3144", "#2274A5", "#32936F",
"#D00000", "#FFBF00", "#8D2AB7", "#27182E", "#232534"]
# the ones that we actually care about, in this case we care about distribution between 0 to 3
bin_size_comp = 30
actual_bin_size_comp = 50 # position 40 and 50 are for outliers and negative intervals
bins_comp = [x/10.0 for x in list(range(0, actual_bin_size_comp+1, 1))]
# < -16, -16, -15, -14, -13, -12, -11, -10, -9, -8, -7, -6, -5, -4, -3, -2, -1, 0, 1, 2, >=3, placeholder, empty
bin_size_raw = 22 # the ones that we actually care about, in this case we care about distribution between 10^-17 to 10^3
actual_bin_size_raw = 23 # position 50 for outliers, position 60 for negatives
bins_raw = list(range(-17, 6, 1))
def plot_raw(datas, test_name, system = "mac"):
plt.rc('xtick', labelsize=7) # fontsize of the tick labels
plt.rc('ytick', labelsize=10) # fontsize of the tick labels
plt.rc('legend', fontsize=10) # fontsize of the tick labels
bar_width = 0.5/(len(datas))
all_items = []
for item in datas:
all_items += datas[item]
average = statistics.median(all_items)
if average <= 0 :
average = sum(all_items) * 1.0 / len(all_items)
if average <=0:
average = math.pow(10, -16)
average_log = int(math.floor(math.log10(average)))
limit_down = average_log - 4
limit_up = average_log + 4
count = 0
bins = list(range(21))
bins = [x * 0.5 + limit_down - 1 for x in bins]
# print(bins)
# print(len(bins))
negative_count = 0
ax = plt.axes()
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
plt.xlim([limit_down - 1, limit_up + 2.5])
plt.ylim([0, 100])
plt.xticks([limit_down - 1] + list(range(limit_down, limit_up + 2)), [r'$\leq10^{' + str(limit_down) + r'}$'] + [r'$10^{' + str(x) + r'}$' for x in list(range(limit_down, limit_up))] + [r'$\geq10^{' + str(limit_up) + r'}$', "EMPTY"])
for item in datas:
belows = 0
ranges = [0] * 16
aboves = 0
negatives = 0
for gap in datas[item]:
if gap < 0:
negatives += 1
elif gap == 0:
belows += 1
else:
## compute the log value
log_value = math.log10(gap)
if log_value <= limit_down:
belows += 1
elif log_value >= limit_up:
aboves += 1
else:
position = int((log_value - limit_down) * 2)
ranges[position] += 1
belows = belows * 100.0 / len(datas[item])
ranges = [x * 100.0 / len(datas[item]) for x in ranges]
aboves = aboves * 100.0 / len(datas[item])
negatives = negatives * 100.0 / len(datas[item])
# print(belows)
# print(ranges)
# print(aboves)
# print(negatives)
bar = plt.bar([x+bar_width / 2 + bar_width * count for x in bins], [0] * 2 + ranges + [0] * 3, label=item, color=colors[count], alpha=1.0, width=bar_width)
bar = plt.bar([x+bar_width * 1 + bar_width * 2 * count for x in bins], [belows, 0] + [0] * 16 + [aboves, 0] + [negatives] , color=colors[count], alpha=1.0, width=bar_width *2)
if (negatives > 0):
rect = bar[-1]
plt.text(limit_up + 1.5, rect.get_height(), ('%s:\n%f' % (item, negatives)).rstrip('0').rstrip('.')+"%"+"\n\n"*negative_count, ha='center', va='bottom')
negative_count += 1
count += 1
plt.legend(prop={'size': 6})
plt.title(test_name)
# plt.show()
save_plot_name = "graphs/raw/" + test_name + "_gap_raw_" + system + ".pdf"
plt.savefig(save_plot_name, bbox_inches='tight',
pad_inches=0, dpi=200)
plt.close()
print(test_name + " raw interval size graph generated")
def plot_comp(datas, test_name, system = "mac"):
pass
def compute_comparison(base, num):
frac = 1.0
if (base == 0 and num == 0):
frac = 1
elif (base == 0 and num != 0):
frac = 999
elif (base != 0 and num == 0):
frac = 0
else:
frac = float(num / base)
return frac
def plot_data(file_name, system = "mac"):
f = open(file_name, 'r')
last_test_name = None
datas = {}
numerator = 1
denominator = 1
for line in f:
splitted = line.strip().split(", ")
test_name = ", ".join(splitted[1:-2])
method_name = splitted[-2]
if last_test_name != test_name:
if last_test_name!=None:
plot_raw(datas, last_test_name, system)
plot_comp(datas, last_test_name, system)
## clear data
datas = {}
## set the new test name
last_test_name = test_name
method_names.clear()
## put method name in dictionary
if not method_name in method_names:
method_names.append(method_name)
datas[method_name] = []
## get the gap size in numerator and denominator
if (len(splitted[-1].split("/")) == 2):
numerator = int(splitted[-1].split("/")[0])
denominator = int(splitted[-1].split("/")[1])
else:
numerator = int(splitted[-1].split("/")[0])
denominator = 1
datas[method_name].append(float(Fraction(numerator, denominator)))
plot_raw(datas, last_test_name, system)
plot_comp(datas, last_test_name, system)
def main():
plot_data("build/gaps.txt", str(platform.platform()))
if __name__ == "__main__":
main()