TETRIS-code/MakePercentilePlot.py at master · skalyanb/TETRIS-code · 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
import os
import sys
import time
import pandas as pd
import matplotlib.pyplot as plt

dirname = os.path.dirname
sys.path.append(dirname(dirname(os.path.realpath(__file__))))

def skip_rows (index):
    if index <4 or index > 5:
        return True
    return False

def plot_estimates (data_dir, out_filename, x_max):
    X=[]
    Y=[]
    max_err_percent = 10
    min_err_percent = 10

    max_band_percent = 1.5
    min_band_percent = 1.5

    for i, file in enumerate(os.listdir(data_dir)):
        path = os.path.join(data_dir, file)
        if os.path.isdir(path):
            # skip directories
            continue
        if not file.startswith('.'):
            data_file = data_dir + file
            df_tri = pd.read_csv(data_file, sep=",",skiprows=lambda x: skip_rows(x))
            triangle_count = df_tri["triangles"][0]
            df_rawdata = pd.read_csv(data_file, sep=",",skiprows=15)
            # keep only the data that are withing 80 percentile
            upper = df_rawdata["triangle_estimate"].quantile(.75)
            lower = df_rawdata["triangle_estimate"].quantile(.25)
            df_rawdata = df_rawdata[df_rawdata["triangle_estimate"] < upper]
            df_rawdata = df_rawdata[df_rawdata["triangle_estimate"] > lower]
            Y.extend(df_rawdata["triangle_estimate"])
            X.extend(df_rawdata["fraction_of_edges_seen"])

    fig, ax = plt.subplots()

    y_band_min = triangle_count*(1-min_band_percent/100.0)
    y_band_max = triangle_count*(1+max_band_percent/100.0)
    ax.axhline(y=triangle_count, linewidth=2, color='r', linestyle='-',label='Count')
    ax.axhline(y=y_band_min, color='r', linestyle='--')
    ax.axhline(y=y_band_max, color='r', linestyle='--')

    #scatter plot
    ax.scatter(X, Y, s=20, c='blue', marker='o')

    #change axes ranges
    y_min = triangle_count * (1-min_err_percent/100.0)
    y_max = triangle_count * (1+max_err_percent/100.0)
    ax.set_ylim(y_min, y_max)

    x_min = 0.0
    x_max = x_max
    ax.set_xlim(x_min,x_max)

    relative_pos_band_exact = 1.0*min_err_percent / (min_err_percent + max_err_percent)
    relative_pos_band_min = 1.0* (min_err_percent-min_band_percent) / (min_err_percent + max_err_percent)
    relative_pos_band_max = 1.0* (min_err_percent+max_band_percent) / (min_err_percent + max_err_percent)

    ax.text(1.02, relative_pos_band_exact, '0%',
            horizontalalignment='left',
            verticalalignment='center',
            #rotation='vertical',
            transform=ax.transAxes
            )
    ax.text(1.02, relative_pos_band_max, '+'+str(max_band_percent)+'%',
            horizontalalignment='left',
            verticalalignment='center',
            #rotation='vertical',
            transform=ax.transAxes
            )
    ax.text(1.02, relative_pos_band_min, '-'+str(min_band_percent)+'%',
            horizontalalignment='left',
            verticalalignment='center',
            #rotation='vertical',
            transform=ax.transAxes
            )
    ax.text(1.02, 1, '+'+str(max_err_percent)+'%',
            horizontalalignment='left',
            verticalalignment='center',
            #rotation='vertical',
            transform=ax.transAxes
            )
    ax.text(1.02, 0, '-'+str(min_err_percent)+'%',
            horizontalalignment='left',
            verticalalignment='center',
            #rotation='vertical',
            transform=ax.transAxes
            )


    #add title
    ax.set_title('Accuracy vs Observed Graph')

    #add x and y labels
    ax.set_xlabel('Percentage of Edges Visited')
    ax.set_ylabel('Triangle Estimates')

    #show plot
    #plt.show()
    timestr = time.strftime("%Y%m%d-%H%M%S")

    fig.savefig("output/plots/variance/small_fraction/"+out_filename+"-"+timestr+".eps",format='eps')


if __name__ == "__main__":

    # file_name = "soc-flickr-und"
    # x_max = 1.1


    file_name = "soc-sinaweibo"
    x_max = 1.1

    # file_name = "soc-flickr"
    # x_max = 1.1

    # file_name = "soc-twitter-konect"
    # x_max = 0.11

    # file_name = "soc-livejournal"
    # x_max = 1.1

    # file_name = "soc-orkut"
    # x_max = 1.1

    # file_name = "socfb-A-anon"
    # x_max = 1.1

    data_dir = "output/plot_data/variance_plot_data/"+file_name+".edges/EstTriByRWandWghtedSampling/small_fraction/"
    out_filename = file_name

    plot_estimates(data_dir,out_filename,x_max)

    #plot_comparison(data_dir_1,data_dir_2)